Variations in Oncology Consultations: How Dictation Allows Variations to be Documented in Standardized Ways

Physicians' and Patients' Expectations From Digital Agents for Consultations: Interview Study Among Physicians and Patients

BACKGROUND

Physicians are currently overwhelmed by administrative tasks and spend very little time in consultations with patients, which hampers health literacy, shared decision-making, and treatment adherence.

OBJECTIVE

This study aims to examine whether digital agents constructed using fast-evolving generative artificial intelligence, such as ChatGPT, have the potential to improve consultations, adherence to treatment, and health literacy. We interviewed patients and physicians to obtain their opinions about 3 digital agents-a silent digital expert, a communicative digital expert, and a digital companion (DC).

METHODS

We conducted in-depth interviews with 25 patients and 22 physicians from a purposeful sample, with the patients having a wide age range and coming from different educational backgrounds and the physicians having different medical specialties. Transcripts of the interviews were deductively coded using MAXQDA (VERBI Software GmbH) and then summarized according to code and interview before being clustered for interpretation.

RESULTS

Statements from patients and physicians were categorized according to three consultation phases: (1) silent and communicative digital experts that are part of the consultation, (2) digital experts that hand over to a DC, and (3) DCs that support patients in the period between consultations. Overall, patients and physicians were open to these forms of digital support but had reservations about all 3 agents.

CONCLUSIONS

Ultimately, we derived 9 requirements for designing digital agents to support consultations, treatment adherence, and health literacy based on the literature and our qualitative findings.

“Truly holistic?” Differences in documenting physical and psychosocial needs and hope in Portuguese palliative patients

Objective

Palliative care (PC) aims to improve patients' and families' quality of life through an approach that relieves physical, psychosocial, and spiritual suffering, although the latter continues to be under-assessed and under-treated. This study aimed to describe the prevalence of physical, psychosocial, and hope assessments documented by a PC team in the first PC consultation.

Method

The retrospective descriptive analysis of all first PC consultations registered in our anonymized database (December 2018–January 2020), searching for written documentation regarding (1) Edmonton Symptom Assessment Scale (ESAS) physical subscale (pain, tiredness, nausea, drowsiness, appetite, shortness of breath, constipation, insomnia, and well-being), (2) the single question “Are you depressed?” (SQD), (3) the question “Do you feel anxious?” (SQA), (4) feeling a burden, (5) hope-related concerns, (6) the dignity question (DQ), and (7) will to live (WtL).

Results

Of the 174 total of patients anonymously registered in our database, 141 PC home patients were considered for analysis; 63% were male, average age was 70 years, the majority had malignancies (82%), with a mean performance status of 52%. Evidence of written documentation was (1) ESAS pain (96%), tiredness (89%), nausea (89%), drowsiness (79%), appetite (89%), shortness of breath (82%), constipation (74%), insomnia (72%), and well-being (52%); (2) the SQD (39%); (3) the SQA (11%); (4) burden (26%); (5) hope (11%); (6) the DQ (33%); and (7) WtL (33%).

Significant differences were found between the frequencies of all documented items of the ESAS physical subscale (29%), and all documented psychosocial items (SQD + SQA + burden + DQ) (1%), hope (11%), and WtL (33%) (p = 0.0000; p = 0.0005; p = 0.0181, respectively).

Significance of results

There were differences between documentation of psychosocial, hope, and physical assessments after the first PC consultation, with the latter being much more frequent. Further research using multicenter data is now required to help identify barriers in assessing and documenting non-physical domains of end-of-life experience.

Speech recognition for clinical documentation from 1990 to 2018: a systematic review

OBJECTIVE

The study sought to review recent literature regarding use of speech recognition (SR) technology for clinical documentation and to understand the impact of SR on document accuracy, provider efficiency, institutional cost, and more.

MATERIALS AND METHODS

We searched 10 scientific and medical literature databases to find articles about clinician use of SR for documentation published between January 1, 1990, and October 15, 2018. We annotated included articles with their research topic(s), medical domain(s), and SR system(s) evaluated and analyzed the results.

RESULTS

One hundred twenty-two articles were included. Forty-eight (39.3%) involved the radiology department exclusively and 10 (8.2%) involved emergency medicine; 10 (8.2%) mentioned multiple departments. Forty-eight (39.3%) articles studied productivity; 20 (16.4%) studied the effect of SR on documentation time, with mixed findings. Decreased turnaround time was reported in all 19 (15.6%) studies in which it was evaluated. Twenty-nine (23.8%) studies conducted error analyses, though various evaluation metrics were used. Reported percentage of documents with errors ranged from 4.8% to 71%; reported word error rates ranged from 7.4% to 38.7%. Seven (5.7%) studies assessed documentation-associated costs; 5 reported decreases and 2 reported increases. Many studies (44.3%) used products by Nuance Communications. Other vendors included IBM (9.0%) and Philips (6.6%); 7 (5.7%) used self-developed systems.

CONCLUSION

Despite widespread use of SR for clinical documentation, research on this topic remains largely heterogeneous, often using different evaluation metrics with mixed findings. Further, that SR-assisted documentation has become increasingly common in clinical settings beyond radiology warrants further investigation of its use and effectiveness in these settings.