To understand the effectiveness of AMC as an innovative method of health service delivery, we conducted a study in collaboration with the UFH group.
Research method
In this study, 200 patients were randomly selected from the outpatient department of UFH, half of the patients used AMC-related telemedicine consultation and follow-up processes (AMC group), and the other half usually (UC group)), including in-person outpatient visits and on-site follow-up. An online survey on treatment effectiveness based on waiting time and patient satisfaction was conducted for AMC and UC groups on our telemedicine mobile platform. Specifically, we developed a survey on patient health treatment satisfaction (PSQ18 Questionnaire 13) and patient willingness to use AMC.
This study was designed to be a retrospective observational design 14 that included existing patients in UFH hospitals and clinics (Beijing, Shanghai, Qingdao, Guangzhou, and Shenzhen). AMC Group’s process consists of telemedicine consultation, mobile heart rate, body temperature detection, mobile diagnostics, and mobile follow-up. In contrast, the UC group process involves a direct visit to the hospital for treatment of the relevant disease.
We prepared an AMC mock room for telemedicine consultations using pad screens 4g/5g or clinic Wi-Fi setups, along with wearable monitoring devices such as heart rate and body temperature detection. Nurses will also sit in a mock room and help patients complete the assessment process, including self-tests of heart rate and body temperature using wearable monitoring devices.
In this study, the study follow-up period was set at 3 months, and participants in this study provided the results of the above survey and objective process 3 months after treatment introduction. Please note that patients younger than 18 years of age and those unwilling to participate in medically related research were excluded from this study based on the methodology presented in 15 .
Regarding sample size selection, we are piloting with 200 patients with an expected effect size of 1.5, so the minimum study sample size was 152 to meet the statistical power requirements of a = 0.05 and 80%. Become a human subject. Conversely, the sample size of 200 that we tend to include in this study will perform 90% of the statistical power from the previous settings in terms of error bars and confidence intervals. The sample size calculation tool used was Microsoft Sample Size Calculator TM16. The sample size calculation should follow the statistical formula: n = (Zα/2 + Zβ)2 *2*σ2 / d2. Zα/2 is the critical value in the α/2 normal distribution with a 95% confidence interval, while Zβ is the critical value at the β 0.2 level representing 80% power, and σ2 is the population variance, d is the expected difference 17.
Research results
Table 1 Subject demographics and characteristics.
Table 1 summarizes the demographics of the study. The age of the participants was 18% under 30 years old, 23% between 31 and 40 years old, 21% between 41 and 50 years old, 28% between 51 and 60 years old, and 10% over 60 years old. The mean age of the subjects was 44 years and the median age was 45 years. 52% of study subjects were male and 48% female. 26% of patients had a graduate degree, 33% had an undergraduate degree, 21% had a college level, 14% had a high school level, and 6% had an elementary level. Socio-economic sataus is classified as annual income in US dollars. 20% of patients have an annual income of 200,000 or more, 39% have an annual income of between 100,000 and 200,000, 23% have an annual income of between 50,000 and 100,000, 13% have an annual income of between 20,000 and 50,000, and only a few have an annual income of less than 50,000. It was 5%.
Figure 3
Treatment effects on different waiting times: AMC and usual care.
As shown in Figure 3, the X-axis is the difference in treatment processing time (both AMC and UC groups), and the Y-axis is the patient-reported treatment effect (from a telemedicine mobile app survey). The blue line is the expected treatment effect of patients for the AMC group with different treatment times, and the orange line is the respective result for the UC group.
The results show that as processing time increases, the therapeutic effect expected by patients decreases. From the blue line, when treatment processing time or waiting time exceeds 52 minutes, the patient’s expected AMC treatment effect decreases to less than 50%. For the UC group, 50% of the treatment effect patients expect comes from just 30 minutes of waiting time.
When using AMC, compared to usual care, patient tolerance for a 50% treatment effect is greater than 20 minutes. On the other hand, the results show that AMC has a higher therapeutic effect than ever that patients expect compared to Usual Care. However, we see that the difference between AMC and usual care narrows when treatment processing time exceeds 60 minutes. In summary, AMC has much higher therapeutic efficacy compared to usual care for various health conditions with different treatment times.
The next study we conducted was the association between patient health treatment satisfaction (PSQ18 questionnaire) and patient willingness to use AMC. The results are summarized in Figure 4.
Figure 4
Association between patient health treatment satisfaction (PSQ18 questionnaire) and patient willingness to use AMC.
Because PSQ-18 score and patient willingness to use AMC are both numerical variables, we performed a simple regression analysis to assess the association.
From the figure above, we can see that patient satisfaction treatment process (using the WHO validated PSQ18 questionnaire) is positively correlated with patient willingness to use AMC. This result indicates that when the probability that a patient wishes to use AMC increases, the patient’s satisfaction with the treatment process also increases. Similarly, when AMC motivation decreases, so does patient satisfaction. With a common cutoff of a patient satisfaction score of 80%, we found that these subjects also had an approximately 80% probability of using AMC as a healthcare treatment process. Alternatively, such results reflect that AMC not only reduces processing and waiting times for healthcare access, but also increases patient satisfaction.
Table 2. Associations of patient satisfaction with multivariate regression analysis.
We also perform multivariate regression analysis to examine the association between patient satisfaction and AMC willingness utilization and other covariates and confounders such as education level, socio-economic status, age, and gender. . Educational levels are divided into five classes: graduate, undergraduate, university, and high school to elementary school levels. Socioeconomic status is classified as annual income in US dollars.
The results in Table 2 showed that the willingness to use AMC was positively related to patient satisfaction (coefficient 77.78, P value < 0.01). When patients demonstrate that they are enthusiastic about using autonomous mobile clinic visits, they receive higher patient satisfaction scores. On the other hand, age is negatively related to patient satisfaction. (coefficient − 0.41, P value 0.01). Education level and socio-economic status were positively correlated with patient satisfaction with coefficient 8.83, P value < 0.01 and coefficient 6.96, P value < 0.01, respectively. All factor results are adjusted for other covariates. (e.g., AMC willingness results are adjusted for age, gender, education level, and socio-economic status; results for other factors are adjusted for each).
