Search

Many patients have difficulty communicating with their healthcare providers (Aldoory, 2017; Brooks et al., 2020; Watson, 2019). Teach-Back is an evidence-based health literary communication tool that is useful for ensuring patients understand the information they receive; with evidence after exposure to Teach-Back, patients exhibit increased disease-specific knowledge, adherence, self-efficacy, and improved health outcomes (Hong et al., 2020a; Talevski et al., 2020). Healthy People 2030 endorsed Teach-Back as an intervention for improving patient comprehension of health information in clinical settings (ODPHP, 2020). However, Teach-Back has not been widely implemented in these settings (Brooks et al., 2020). Not much is known regarding the comprehensive factors which predict clinicians’ use of Teach-Back. This study is a follow-up to Feinberg et al. (2019) which found medical residents increased their use of Teach-Back after attending an educational presentation. Variables not examined in the original analysis are explored in the following research questions: After an educational intervention, is there a significant relationship between patients’ exposure to Teach-Back by medical residents and patients’ highest educational level, perceived health status, diagnoses, reason for visit, new or returning patient status, gender and/or the conversation length during consultation? What are the relationships between the total number of times Teach-Back is used post-intervention and the medical resident’s age, gender, race, or main language? Bivariate and Kendall’s tau correlations revealed new or returning patient status (χ2 = 5.430, p < .05) and conversation length (rt = .307, p < .01), respectively, were significantly associated with patient’s Teach-Back exposure after the residents’ participation in the training intervention. The subsequent binary logistic regression revealed only conversation length (B = .061, p < .05) was a significant predictor of Teach-Back exposure. No statistically significant relationships between the total number of times Teach-Back was used post-intervention and medical residents’ personal characteristics were found. Findings add to original study results of factors which facilitate Teach-Back exposure. Longer consultations and new patient visits were more likely to involve Teach-Back use than other factors. Given the benefits of Teach-Back for all patients, a universal precautions approach is recommended to facilitate its widespread adoption in clinical settings.

The treatment landscape of multiple myeloma (MM) had significantly evolved in the last decade. Treatment duration can be influenced by several disease-specific and patient-related features, and accurately predicting when a patient will require their next line of treatment is challenging. This study evaluated and compared prediction models for a second line treatment among newly diagnosed MM patients using machine learning (ML) algorithms. Non-stem cell transplant adult patients (≥ 65 years old) with MM diagnosis in a large national administrative claims database between 1/1/2013 and 8/31/2017 were identified. Health plan enrollment of 6 months prior and ≥ 1 month after index diagnosis, and ≥ 1 MM-related pharmacy claim after index diagnosis were required. Time on first treatment was assessed if a patient received a second line treatment within 12 months. Statistical models were logistic regression and Cox proportional hazards regression. ML classification models, such as logistic regression (LR), penalized regression (LASSO, ridge, and elastic net), random forests (RF), and survival-based ML models, such as random survival forests (RSF) were used to develop prediction models. Models were trained using 70% of the randomly split data. A 10-fold cross-validation repeated 3 times and down-sampling method were used for training classification models. Classification and survival-based ML model performances were evaluated using area under the curve (AUC) with confusion matrix and Harrell’s concordance error rate, respectively. This study identified 1,285 eligible MM patients, and 333 (26%) of those had a second line treatment within 12 months. Statistical logistic model had better performance (AUC= 0.75) than RF. However, RF achieved the best performance with an AUC= 0.71 (accuracy= 67%, recall= 66%, F1= 51%, and F2= 59%) among other classification models. Other classification models were comparable (AUC= 0.68-0.69, accuracy= 64%-67%, recall= 59%-61%, F1= 46%-48%, F2= 52%-55%). In survival-based models, the Cox model performed better (error rate= 0.302) than ML. The log-rank splitting approach gave the best RSF performance (error rate= 0.335). In classification methods, RF and statistical logistic models both achieved good performance. However, in survival-based methods, the Cox model had better performance than RSF. The use of these methods may be subject to the study needs.

Objectives: Palliative care (PC) services and patients differ across countries. Data on PC delivery paired with medical and self-reported data are seldom reported. Aims were to describe (1) PC organisation and services in participating centres and (2) characteristics of patients in PC programmes. Methods: This was an international prospective multicentre study with a single web-based survey on PC organisation, services and academics and patients' self-reported symptoms collected at baseline and monthly thereafter, with concurrent registrations of medical data by healthcare providers. Participants were patients ≥18 enrolled in a PC programme. Results: 30 centres in 12 countries participated; 24 hospitals, 4 hospices, 1 nursing home, 1 home-care service. 22 centres (73%) had PC in-house teams and inpatient and outpatient services. 20 centres (67%) had integral chemotherapy/radiotherapy services, and most (28/30) had access to general medical or oncology inpatient units. Physicians or nurses were present 24 hours/7 days in 50% and 60% of centres, respectively. 50 centres (50%) had professorships, and 12 centres (40%) had full-time/part-time research staff. Data were available on 1698 patients: 50% females; median age 66 (range 21-97); median Karnofsky score 70 (10-100); 1409 patients (83%) had metastatic/disseminated disease; tiredness and pain in the past 24 hours were most prominent. During follow-up, 1060 patients (62%) died; 450 (44%)

Total knee arthroplasty (TKA) remains the most effective treatment for osteoarthritis of the knee. Multiple factors are involved in the recovery of knee function and the research to date is equivocal on which factors are most influential in the recovery of knee function. The purpose of this study is to identify factors that may influence a person's functional improvement after a TKA. Retrospective data on a convenience sample of 95 subjects were obtained from a registry of patients kept at the Cleveland Center for Joint Reconstruction and Lutheran Hospital, Cleveland, Ohio. The conceptual model for this study was adapted from the Andersen Behavioral Model. Patient characteristics were used to describe the patient population attributes, the utilization of health services was used to describe the health behaviors, and functional improvement was used as a measure for outcomes. A descriptive correlational design was used. The data were analyzed utilizing multiple regression. The findings suggest that the most consistent predictors of functional improvement post-TKA were the preoperative need variables (KSS and SF-12) reflecting physical and mental functioning. Gender, age and BMI were also predictors of functional improvement, though not consistently through-out the analyses. The results suggest that in order to gain a fuller understanding of functional improvement post-TKA, future research should include more specific measures for the evaluation of knee function pre-TKA. The findings from this study are not generalizable due to the lack of variability in the sample. The population was homogenous in that they were relatively healthy and had some form of healthcare coverage. To help facilitate generalizability, further research should incorporate economic and social diversity.

The unplanned hospital readmissions in elderly with heart failure consume a considerable amount of healthcare and social resources. Thus, knowledge of the characteristics and predictive factors for patients who are at high risk for hospital readmission could provide a better understanding upon which to develop effective interventions aimed at improving patient outcomes in heart failure patients. The purpose of this study was to examine the patient characteristics and factors related to hospital readmission in heart failure patients. Multivariate analyses of 98heart failure patients treated and discharged during May 1, 2006 to May, 1, 2007 were performed. Andersen's Behavioral Model of Health Service Utilization was used to organize factors including population characteristics (predisposing characteristics, enabling resources, and need), health behaviors/use of health services (hospital length of stay, home healthcare, CCU stay, device implantation), and outcome (readmission). Differences in population characteristics between readmitted and not readmitted patients were compared. Readmitted study patients had a significantly lower hemoglobin (t = 2.23, p = .03) and significantly higher sodium level (t = -2.3, p = .02). In addition,, patients who were diagnosed with diabetes (χ2 = 4.4, p = .04) or who had physical impairment at discharge (χ2 = 9.2, p = .002) had higher readmission rate than the not readmitted HF patients. Use of health services was compared. The readmitted HF patients were more likely to have a CCU stay at their index admission (26% vs. 9%). There were not significant differences on hospital length of stay, having device implantation, and home healthcare between readmitted and not readmitted patients. A lower LVEF, the presence of physical impairment were associated with longer hospital stays. Heart failure patients with physical impairment and those with diabetes were more likely to received home healthcare before their initial hospitalization. Conclusion: Population characteristic of HF patients were influential on hospital readmission. Among three of the sub-concepts of population characteristics, need had the most predictability in this study, consistent with the theory. However, the moderating effect of use of health service between population characteristics and outcome was not supported in this study.

Due to increased provision by the private sector, the health care industry in New Zealand is rapidly becoming more competitive (Zwier, 1998). As a result, the implementation of reliable service quality questionnaires is becoming more important as greater value is placed on patient satisfaction by health care providers (Zwier, 1998). Despite this, research into health care patients' perceptions of service quality dimensions in New Zealand is scarce. This study takes a step towards addressing this deficiency by identifying patients' perceptions of service quality dimensions in North Canterbury, New Zealand. The findings of this study are based on the analysis of a sample of 389 respondents interviewed by telephone. The findings indicate that the service quality dimensions identified in this study differ in number and content from the widely adopted service quality dimensions identified by Parasuraman et al. (1988). In addition, health care patients in North Canterbury perceive service quality dimensions relating to the core product in health care delivery (for example, outcome and reliability) as more important than service quality dimensions relating to the peripheral product in health care delivery (for example, food, access and tangibles). Finally, the results of this study indicate that patients with different geographic, demographic, psychographic and behaviouristic characteristics have different needs and wants during health care delivery and therefore perceive different service quality dimensions as important. This study contributes to services marketing theory by providing empirical research into health care patients’ perceptions of service quality dimensions in a setting different from Europe and North America. In addition, this study provides health care professionals with insights regarding: the development of more reliable service quality questionnaires, the allocation of health care resources, understanding the needs of out-patients compared to in-patients, and understanding the needs of private patients compared to public patients.