Patients' preferences for attributes related to health care services at hospitals in Amhara Region, northern Ethiopia: a discrete choice experiment

Addressing fairness issues in deep learning-based medical image analysis: a systematic review

Deep learning algorithms have demonstrated remarkable efficacy in various medical image analysis (MedIA) applications. However, recent research highlights a performance disparity in these algorithms when applied to specific subgroups, such as exhibiting poorer predictive performance in elderly females. Addressing this fairness issue has become a collaborative effort involving AI scientists and clinicians seeking to understand its origins and develop solutions for mitigation within MedIA. In this survey, we thoroughly examine the current advancements in addressing fairness issues in MedIA, focusing on methodological approaches. We introduce the basics of group fairness and subsequently categorize studies on fair MedIA into fairness evaluation and unfairness mitigation. Detailed methods employed in these studies are presented too. Our survey concludes with a discussion of existing challenges and opportunities in establishing a fair MedIA and healthcare system. By offering this comprehensive review, we aim to foster a shared understanding of fairness among AI researchers and clinicians, enhance the development of unfairness mitigation methods, and contribute to the creation of an equitable MedIA society.

Patients' Preferences for Antiretroviral Therapy Service in Northwest Ethiopia: A Discrete Choice Experiment

Objective. We aim to evaluate patients' preferences for antiretroviral therapy (ART) to enhance shared decision making in clinical practice in Northwest Ethiopia. Methods. A discrete choice experiment approach was used among adult patients from 36 randomly selected public health facilities from February 6, 2023, to March 29, 2023. A literature review, qualitative work, ranking and rating surveys, and expert consultation were used to identify the attributes. Location, provider, frequency of visit, appointment modality, refill time, and cost of visit were the 6 ART service features chosen. Participants were given the option of choosing between 2 hypothetical differentiated ART delivery models. Mixed logit and latent class analysis were used. Results: Four hundred fifty-six patients completed the choice task. Respondents preferred to receive ART refills alone at health facilities by health care workers without having to have frequent visits and with reduced cost of visit. Overall, the participants valued the cost of the visit the most while they valued the timing of ART refill the least. Participants were willing to pay only for the attributes of frequency of visit and medication refill time. The latent class model with 3 classes provided the best model fit. Location, cost, and frequency were the most important attributes in class 1, class 2, and class 3, respectively. Income and marital status significantly predicted class membership. Conclusions. Respondents preferred to receive refills at health facilities, less frequent visits, individual appointments, service provision by health care workers, and reduced cost of visit. The cost attribute had the greatest impact on the choice of patients. Health care workers should consider the preferences of patients while providing ART services to meet patients' expectations and choices.

Highlights

A discrete choice experiment was used to elicit patient preferences.People living with HIV preferred receiving medication refills at health facilities, less frequent visits, individual appointments, service delivery by health care workers, and lower visit costs.Health care workers should consider the preferences of patients while providing ART service to meet their expectations and choices.Scaling up differentiated HIV treatment services is crucial for patient-centered care.

What influences slum residents' choices of healthcare providers for common illnesses? Findings of a Discrete Choice Experiment in Ibadan, Nigeria

Urban slum residents have access to a broad range of facilities of varying quality. The choices they make can significantly influence their health outcomes. Discrete Choice Experiments (DCEs) are a widely-used health economic methodology for understanding how individuals make trade-offs between attributes of goods or services when choosing between them. We carried out a DCE to understand these trade-offs for residents of an urban slum in Ibadan, Nigeria. We conducted 48 in-depth interviews with slum residents to identify key attributes influencing their decision to access health care. We also developed three symptom scenarios worded to be consistent with, but not pathegonian of, malaria, cholera, and depression. This led to the design of a DCE involving eight attributes with 2-4 levels for each. A D-efficient design was created, and data was collected from 557 residents between May 2021 and July 2021. Conditional-logit models were fitted to these data initially. Mixed logit and latent class models were also fitted to explore preference heterogeneity. Conditional logit results suggested a substantial Willingness-to-pay (WTP) for attributes associated with quality. WTP estimates across scenarios 1/2/3 were N5282 / N6080 / N3715 for the government over private ownership, N2599 / N5827 / N2020 for seeing a doctor rather than an informal provider and N2196 / N5421 /N4987 for full drug availability over none. Mixed logit and latent class models indicated considerable preference heterogeneity, with the latter suggesting a substantial minority valuing private over government facilities. Higher income and educational attainment were predictive of membership of this minority. Our study suggests that slum residents value and are willing to pay for high-quality care regarding staff qualifications and drug availability. It further suggests substantial variation in the perception of private providers. Therefore, improved access to government facilities and initiatives to improve the quality of private providers are complementary strategies for improving overall care received.

Effect of hospital attributes on patient preference among outpatient attendants in Wolaita Zone, Southern Ethiopia: discrete choice experiment study

BACKGROUND

Patient preference has preceded the use of health care services, and it has been affected by different hospital attributes. Meanwhile, the number of patients receiving vital health intervention is particularly low in Ethiopia. Therefore, this study aimed to determine the effect of hospital attributes on patient preference for outpatients in the Wolaita area in September 2020.

METHODS

A discrete choice experimental study was applied to determine the effect of hospital attributes on patient preference with a sample size of 1077. The experimental survey was conducted among outpatient attendants selected through a systematic random sampling approach. Six key attributes (competence of healthcare providers; availability of medical equipment and supplies; cost of service; wait time; distance; and hospital reputation) deduced from various hospital attributes were used to elicit the patient preferences. The data was collected from participants through the Open Data Kit application. A random effect probit model with marginal willingness to pay measure and partially log-likelihood analysis was applied to extract important attributes. We used STATA version 15 software for analysis, and the fitness of the model was verified by the calculated p-value for the Wald chi-square with a cut-point value of 0.05.

RESULT

One thousand forty-five patients who received outpatient care participated in the study. The random effect probit results have shown that all hospital attributes included in the study were significantly valued by patients while choosing the hospital (p-value < 0.001). Meanwhile, based on marginal willingness to pay and partial log-likelihood analysis, the competence of health care providers was identified as the most important attribute followed by the availability of medical equipment and supplies in hospitals.

CONCLUSION AND RECOMMENDATION

The results suggested that the quality of health care providers and availability of medical equipment and supply in hospitals would be primary interventional points for improving the patient preference of hospitals. Assessment, education, and training are recommended for enhancing the quality of health care providers. And stock balance checks, inspections, and accreditation are believed to be valuable for improving the availability of equipment and supply in hospitals.

Eliciting preferences of patients about the quality of hospital services in the west of Iran using discrete choice experiment analysis

Objectives

Knowing about accurate customer expectations is the most important step in defining and delivering high-quality services. This study aimed to evaluate the preferences of patients referring to two hospitals in Kermanshah, Iran.

Method

Discrete choice experiment (DCE) method used to elicit preferences of 328 patients who were admitted in two hospitals of Kermanshah city in the west of Iran. Literature review and experts opinion were used to identify a candidate list of attributes related to the quality of cares in hospitals. The final study attributes were quality of physician care, quality of nursing care, waiting time for admission, cleaning of wards and toilets, and behavior of staff. Experimental design applied to extract choice sets of hospitals. The data was analyzed by a conditional logit regression.

Results

The regression results showed the most important predictors of hospital selection by respondents was the good quality of physician care (aOR: 3.18, 95% CI 2.61, 3.87), followed by friendly behavior of staffs (aOR: 2.03, 95% CI 1.81, 2.27), cleanness of wards and toilet (aOR: 1.61, 95% CI 1.40, 1.85), and finally quality of nursing cares (aOR: 1.13, 95% CI 0.89, 1.44). However, increasing waiting time made disutility in the study participants (aOR: 0.69, 95% CI 0.60, 0.80).

Conclusions

Our study finding emphasized some potential opportunity of quality augmentation in hospital sector by paying attention to different quality attributes including quality of physician, friendly behavior of staffs, cleanness of hospital environment and finally quality of nursing cares. Considering patients preferences in decision making process could lead to substantial satisfaction improvement.

Preferences for Artificial Intelligence Clinicians Before and During the COVID-19 Pandemic: Discrete Choice Experiment and Propensity Score Matching Study

Background

Artificial intelligence (AI) methods can potentially be used to relieve the pressure that the COVID-19 pandemic has exerted on public health. In cases of medical resource shortages caused by the pandemic, changes in people’s preferences for AI clinicians and traditional clinicians are worth exploring.

Objective

We aimed to quantify and compare people’s preferences for AI clinicians and traditional clinicians before and during the COVID-19 pandemic, and to assess whether people’s preferences were affected by the pressure of pandemic.

Methods

We used the propensity score matching method to match two different groups of respondents with similar demographic characteristics. Respondents were recruited in 2017 and 2020. A total of 2048 respondents (2017: n=1520; 2020: n=528) completed the questionnaire and were included in the analysis. Multinomial logit models and latent class models were used to assess people’s preferences for different diagnosis methods.

Results

In total, 84.7% (1115/1317) of respondents in the 2017 group and 91.3% (482/528) of respondents in the 2020 group were confident that AI diagnosis methods would outperform human clinician diagnosis methods in the future. Both groups of matched respondents believed that the most important attribute of diagnosis was accuracy, and they preferred to receive combined diagnoses from both AI and human clinicians (2017: odds ratio [OR] 1.645, 95% CI 1.535-1.763; P<.001; 2020: OR 1.513, 95% CI 1.413-1.621; P<.001; reference: clinician diagnoses). The latent class model identified three classes with different attribute priorities. In class 1, preferences for combined diagnoses and accuracy remained constant in 2017 and 2020, and high accuracy (eg, 100% accuracy in 2017: OR 1.357, 95% CI 1.164-1.581) was preferred. In class 2, the matched data from 2017 were similar to those from 2020; combined diagnoses from both AI and human clinicians (2017: OR 1.204, 95% CI 1.039-1.394; P=.011; 2020: OR 2.009, 95% CI 1.826-2.211; P<.001; reference: clinician diagnoses) and an outpatient waiting time of 20 minutes (2017: OR 1.349, 95% CI 1.065-1.708; P<.001; 2020: OR 1.488, 95% CI 1.287-1.721; P<.001; reference: 0 minutes) were consistently preferred. In class 3, the respondents in the 2017 and 2020 groups preferred different diagnosis methods; respondents in the 2017 group preferred clinician diagnoses, whereas respondents in the 2020 group preferred AI diagnoses. In the latent class, which was stratified according to sex, all male and female respondents in the 2017 and 2020 groups believed that accuracy was the most important attribute of diagnosis.

Conclusions

Individuals’ preferences for receiving clinical diagnoses from AI and human clinicians were generally unaffected by the pandemic. Respondents believed that accuracy and expense were the most important attributes of diagnosis. These findings can be used to guide policies that are relevant to the development of AI-based health care.

Patients' Preferences for Artificial Intelligence Applications Versus Clinicians in Disease Diagnosis During the SARS-CoV-2 Pandemic in China: Discrete Choice Experiment

Background

Misdiagnosis, arbitrary charges, annoying queues, and clinic waiting times among others are long-standing phenomena in the medical industry across the world. These factors can contribute to patient anxiety about misdiagnosis by clinicians. However, with the increasing growth in use of big data in biomedical and health care communities, the performance of artificial intelligence (Al) techniques of diagnosis is improving and can help avoid medical practice errors, including under the current circumstance of COVID-19.

Objective

This study aims to visualize and measure patients’ heterogeneous preferences from various angles of AI diagnosis versus clinicians in the context of the COVID-19 epidemic in China. We also aim to illustrate the different decision-making factors of the latent class of a discrete choice experiment (DCE) and prospects for the application of AI techniques in judgment and management during the pandemic of SARS-CoV-2 and in the future.

Methods

A DCE approach was the main analysis method applied in this paper. Attributes from different dimensions were hypothesized: diagnostic method, outpatient waiting time, diagnosis time, accuracy, follow-up after diagnosis, and diagnostic expense. After that, a questionnaire is formed. With collected data from the DCE questionnaire, we apply Sawtooth software to construct a generalized multinomial logit (GMNL) model, mixed logit model, and latent class model with the data sets. Moreover, we calculate the variables’ coefficients, standard error, P value, and odds ratio (OR) and form a utility report to present the importance and weighted percentage of attributes.

Results

A total of 55.8% of the respondents (428 out of 767) opted for AI diagnosis regardless of the description of the clinicians. In the GMNL model, we found that people prefer the 100% accuracy level the most (OR 4.548, 95% CI 4.048-5.110, P<.001). For the latent class model, the most acceptable model consists of 3 latent classes of respondents. The attributes with the most substantial effects and highest percentage weights are the accuracy (39.29% in general) and expense of diagnosis (21.69% in general), especially the preferences for the diagnosis “accuracy” attribute, which is constant across classes. For class 1 and class 3, people prefer the AI + clinicians method (class 1: OR 1.247, 95% CI 1.036-1.463, P<.001; class 3: OR 1.958, 95% CI 1.769-2.167, P<.001). For class 2, people prefer the AI method (OR 1.546, 95% CI 0.883-2.707, P=.37). The OR of levels of attributes increases with the increase of accuracy across all classes.

Conclusions

Latent class analysis was prominent and useful in quantifying preferences for attributes of diagnosis choice. People’s preferences for the “accuracy” and “diagnostic expenses” attributes are palpable. AI will have a potential market. However, accuracy and diagnosis expenses need to be taken into consideration.

Discrete Choice Experiments in Health Economics: Past, Present and Future

OBJECTIVES

Discrete choice experiments (DCEs) are increasingly advocated as a way to quantify preferences for health. However, increasing support does not necessarily result in increasing quality. Although specific reviews have been conducted in certain contexts, there exists no recent description of the general state of the science of health-related DCEs. The aim of this paper was to update prior reviews (1990-2012), to identify all health-related DCEs and to provide a description of trends, current practice and future challenges.

METHODS

A systematic literature review was conducted to identify health-related empirical DCEs published between 2013 and 2017. The search strategy and data extraction replicated prior reviews to allow the reporting of trends, although additional extraction fields were incorporated.

RESULTS

Of the 7877 abstracts generated, 301 studies met the inclusion criteria and underwent data extraction. In general, the total number of DCEs per year continued to increase, with broader areas of application and increased geographic scope. Studies reported using more sophisticated designs (e.g. D-efficient) with associated software (e.g. Ngene). The trend towards using more sophisticated econometric models also continued. However, many studies presented sophisticated methods with insufficient detail. Qualitative research methods continued to be a popular approach for identifying attributes and levels.

CONCLUSIONS

The use of empirical DCEs in health economics continues to grow. However, inadequate reporting of methodological details inhibits quality assessment. This may reduce decision-makers' confidence in results and their ability to act on the findings. How and when to integrate health-related DCE outcomes into decision-making remains an important area for future research.

Underlying determinants of health provider choice in urban slums: results from a discrete choice experiment in Ahmedabad, India

Background

Severe underutilization of healthcare facilities and lack of timely, affordable and effective access to healthcare services in resource-constrained, bottom of pyramid (BoP) settings are well-known issues, which foster a negative cycle of poor health outcomes, catastrophic health expenditures and poverty. Understanding BoP patients’ healthcare choices is vital to inform policymakers’ effective resource allocation and improve population health and livelihood in these areas. This paper examines the factors affecting the choice of health care provider in low-income settings, specifically the urban slums in India.

Method

A discrete choice experiment was carried out to elicit stated preferences of BoP populations. A total of 100 respondents were sampled using a multi-stage systemic random sampling of urban slums. Attributes were selected based on previous studies in developing countries, findings of a previous exploratory study in the study setting and qualitative interviews. Provider type and cost, distance to the facility, attitude of doctor and staff, appropriateness of care and familiarity with doctor were the attributes included in the study. A random effects logit regression was used to perform the analysis. Interaction effects were included to control for individual characteristics.

Results

The relatively most valued attribute is appropriateness of care (β=3.4213, p = 0.00), followed by familiarity with the doctor (β=2.8497, p = 0.00) and attitude of the doctor and staff towards the patient (β=1.8132, p = 0.00). As expected, respondents prefer shorter distance (β= − 0.0722, p = 0.00) but the relatively low importance of the attribute distance to the facility indicate that respondents are willing to travel longer if any of the other statistically significant attributes are present. Also, significant socioeconomic differences in preferences were observed, especially with regard to the type of provider.

Conclusion

The analyses did not reveal universal preferences for a provider type, but overall the traditional provider type is not well accepted. It also became evident that respondents valued appropriateness of care above other attributes. Despite the study limitations, the results have broader policy implications in the context of Indian government’s attempts to reduce high healthcare out-of-pocket expenditures and provide universal health coverage for its population. The government’s attempt to emphasize the focus on traditional providers should be carefully reconsidered.