These authors point out that patient narratives are a potentially valuable but largely unscrutinized source of information. Patients often write free text comments on these surveys, but in the past these comments have been difficult and time-consuming to analyze. These authors used natural language processing via machine learning algorithms to explore the content of negative comments after total shoulder arthroplasty (TSA), their associated factors, and their relationship with traditional measures of patient satisfaction and with perioperative outcomes in 186 patients who had undergone elective primary TSA.
Satisfaction data included patient comments and the Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) survey.
Using a machine-learning-based natural language processing approach, all patient comments were mined for sentiment and classified as positive, negative, mixed, or neutral. Negative comments were further classified into themes.
Most patients (71%) provided at least 1 comment; 32% of the comments were negative, 62% were positive, 5% were mixed, and 1% were neutral. The themes of the negative comments were room condition (27%), time management (17%), inefficient communication (13%), lack of compassion (12%), difficult intravenous (IV) insertion (10%), food (10%), medication side effects (6%), discharge instructions (4%), and pain management (2%). Patients who made negative comments were more likely to be dissatisfied with overall hospital care and with pain management.
Women and sicker patients were more likely to provide negative comments.
Comment: As physicians, we seek to provide not only the safest and most effective patient care, but also a process of care that meets patient expectations. It is surely the case that some patients with superb functional outcomes after shoulder arthroplasty cannot get over their recollection that their hospital room was not kept clean. In a sense we're like a restaurant that wants to provide not only good food, but also a positive dining experience.
Some of the "dissatisfiers" may cost hospitals money to address: single larger rooms are more expensive than shared rooms, spending money on increased staffing can improve responsiveness and allow more time for staff to listen to patient concerns and for effective communication, investing in upgraded food series can assure prompt delivery of warm food to the bedside. Hospitals need to evaluate the cost-benefit of these investments. It also seems the case that these data support efforts to shorten the hospital stay, i.e. getting the patient back to their own home and their own kitchen.
We were particularly interest in the use of machine learning in this study. Efforts to measure patient satisfaction with the process of care traditionally rely on structured tools, such as the Hospital Consumer Assessment of Healthcare Providers and Systems survey, that ask specific, closed-ended questions that may not fully capture the spectrum of patient experiences. Machine learning can enable the efficient mining of free text to extract important themes from the text.
===
We have a new set of shoulder youtubes about the shoulder, check them out at this link.
Be sure to visit "Ream and Run - the state of the art" regarding this radically conservative approach to shoulder arthritis at this link and this link
Use the "Search" box to the right to find other topics of interest to you.
You may be interested in some of our most visited web pages arthritis, total shoulder, ream and run, reverse total shoulder, CTA arthroplasty, and rotator cuff surgery as well as the 'ream and run essentials'
