Forum Replies Created
-
Posts
-
-
Item number 10, page 342
If you reject the test hypothesis because p< 0.05, the chance you are in error is 5% –this sentence is wrong
In my own word; p <0.05 means that the probability of observed this result if the null hypothesis is true is less than 0.05. This probability is low enough to reject the null hypothesis.
Then, if we reject the null hypothesis and accept the alternative hypothesis, the chance of being an error in your decision is 100% because you decide to choose only one hypothesis. -
1. Introduce yourself about your background, what kind of work you are doing that related to statistics.
I am a physician, worked for the university hospital. My clinical work related to the statistic very much in applying the research outcomes to treat the specific patient. I have to read up-to-date experimental studies and judge their results. Is it worth for application to my patient?2. Have you ever learned or applied statistics in your work related to data analysis or statistical analysis? Please share your experience.
I have learned the basic statistical analysis since I was a medical student. I know the basic concepts. However, my statistical analysis skills started 3 years later, during residency training. I have to complete 2 research to be certified as an endocrinologist. I was mostly involved in the enrollment and data collection process but little involved in the statistical analysis method. I have an idea for data manipulation to conclude, but the research assistance helps me do the statistical model and analysis of the data using the SPSS program. -
-
-
In my opinion, replacing old technology with the new one like an updated version of programs. If it claims to be useful as the old one and easier to use this might not be required the TAM assessment. However, if it is a major revision such as change into a different platform or device, require more input information, or make a big change in their workflow. It might need the TAM assessment again.
-
I think one of the external variables that influence the individual perceived usefulness and ease of use is the result of the program that decrease their workload or work redundancy.
For example the EMR system in my hospital, nowadays we do not have a fully EMR system, we still use the paper-based medical record and make a photocopy of it as file storage in the EMR system. Some physicians feel uncomfortable with this new system because it might interrupt the eye-contact with their patients. However, for me, I appreciated this system so much because it is easier to review the patient’s previous medical record (especially for the case with a huge record) and it can solve the problem of missed or disappeared some pages of the medical record. Also, the system provides the medication history that can easily review, and it can be preprinted on the prescription order, this decreases the redundancy in rewrite the medication list on paper-based medical records and prescription orders. -
According to this literature “Burches E, Burches M (2020) Efficacy, Effectiveness and Efficiency in the Health Care: The Need for an Agreement to Clarify its Meaning. Int Arch Public Health Community Med 4:035. doi.org/10.23937/2643-4512/1710035”
Efficacy is the capacity for beneficial change (or therapeutic effect) of a given intervention under ideal or controlled conditions. I think the efficacy is the outcome measurement of a randomized controlled trial.Effectiveness is doing “the right” things under ordinary circumstances to achieve the desired effect. Effectiveness links to the notion of external validity, in that it refers to patients who are visited by physicians in their everyday practice. I think the best example of effectiveness is the outcome of the intervention in the real-world setting, such as the study of the effect of new antidiabetic drugs on HbA1C level reduction in diabetic patients in the clinics. It can be evaluated through observational studies of real practice.
Efficiency is the ratio of the output to the inputs of any system. The input to the system can be physical inputs or financial inputs. While the output in the health sector divided into health services (visits, drugs, admissions) and health outcomes (by way of example: Preventable deaths, functional status, clinical outcomes such as blood pressure or blood sugar control).
Efficiency measures must also explicitly identify the inputs that are used to produce the output of interest. The study design to measure the efficiency should be a cross-sectional study. -
In the literature review related to CDSS I found one interesting publication of cohort study design as described below.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4601128/Title: A two-stage clinical decision support system for early recognition and stratification of patients with sepsis: an observational cohort study
Main objectives of the study: To examined the program’s accuracy in identifying patients at risk of sepsis, the performance of clinical processes, and clinical outcomes.
Main exposure variable of interest: patients got an alert from CDSS as “severe sepsis alert”
Main outcome variable of interest: patients got a confirmed diagnosis from the provider “suspected infection” and an order from the provider in a manner that implies a response to sepsis as “Microbiology culture and IV/PO antibiotics”
Limitations to this study:
1. The setting was a single centre, which may not be generalisable to other clinical settings also the rate of adoption of CDSS may be different.
2. The programme’s adoption by providers may not be fully known because the study began a few weeks after the CDS go-live date; some variance in usability and fidelity may exist because the sepsis programme enabled by the two-stage CDS was relatively new to providers.
3. the study design incorporated a retrospective analysis of cohort data after the launch of the sepsis programme, which may have introduced some selection bias
4. Although the sepsis programme is grounded in current guidelines. However, the guidelines evolve over time and the sepsis programme should evolve too. -
The influencing factor on adoption and use of guideline-based clinical decision support systems in
preventive care into clinical practice.Objectives
To evaluate the factors that influence the adoption and use of guideline-based clinical decision support systems in preventive care into clinical practice; in terms of effort expectancy, performance expectancy, and facilitating
condition. -
Title:
Intention to adopt clinical decision support systems in a developing country: effect of physician’s perceived professional autonomy, involvement and belief: a cross-sectional study.
https://bmcmedinformdecismak.biomedcentral.com/articles/10.1186/1472-6947-12-142Main study objective:
To determine the factors that influence adoption and therefore, use of clinical decision support systems by physicians in hospitals.The sampling method used in this study:
The sampling method used was stratified random sampling.
The physicians were stratified based on the specialty. The size of the hospital (in terms of the total number of physicians) was used to determine the number of physicians to be sampled from a particular hospital. At the next level, within each hospital, the number of physicians sampled in each specialty was determined by the ratio of physicians in that specialty (department) to the total number of physicians in that hospital.Limitation of this study:
1. The generalizability of study results limited to the hospitals around KL, not to all hospitals in Malaysia. because of the advancement of technology in each area is different.
2. The results indicate the factors that influence the adoption of CDSS in Malaysia, not to all developing countries. -
I think one of the confounders in the association between age and the use of contact tracing application is IT proficiency. Young people were born in the digital era. They are familiar with the smartphone. The use of applications does not bother them. In comparison, a lot of elderly feel difficult to use the mobile application.
Also, IT proficiency has a causal relationship with the use of contact tracing applications. Furthermore, IT proficiency is not in between the age and the use of contact tracing application. -
Research title:factors that impact the adoption of the clinical decision support systems in health care setting
Sex- categorical variables— summarized as proportion
Age- continuous variable – summarized as mean(SD)
Clinical experience – continuous variable – summarized as mean(SD)
Specialty area – categorical variables— summarized as proportion (percentage)
Past experience in using CDSS – categorical variables— summarized as proportion (prevalence) -
-
1. How can Blockchain technology provide opportunities for health care?
Blockchain technology uses a decentralized approach that allows the information to be distributed and that each piece of distributed information or commonly known as data, has shared ownership.
The keyword “decentralization” makes this technology have a lot of opportunities in the health care sector because the data can be shared with other people in the chain with, theoretically, enough security because of the encryption of data. For example, the personnel health record, doctor’s consultation network across the world. Nowadays patient health records are kept by each hospital and cannot be shared with other hospitals, even inside or outside the country. This barrier makes patients have an unnecessary investigation when they encounter a new hospital with the same health problem. Furthermore, an expert consultant outside the country can’t be possible because the consultant cannot get access to the full EMR. I think blockchain technology can help in this problem for the health care sector2. According to this study, Blockchain technology presents numerous opportunities for health care. There are any opportunities or potential for applying/implementation blockchain technology to health care in Thailand?
– If there are, what the health care area should be considered?
– If there are not, what the obstacle?I think the area that blockchain technology can apply in Thailand is the remote consultation. Nowadays expert consultation occurs based on the storytelling of the consultee physician, the consultant can not access to the patient’s EMR directly. The recommendation from the consultant is usually not written in the medical record. I think the consultation system with blockchain technology, which consists of the shared patient’s health record from the consultee to the consultant, that consultant physician in that chain can get access to the EMR. Then the official consultation and the official recommendation by an expert, based on the accessible patient’s health record, can be applied.
-
For the public’s opinion in health data storage and sharing during covid situation. I think it changes overtime.
Most of Thai people welcome and allow private data sharing in the very early phase of covid situation.In the context of disease prevention and tracking, like their location, their activities and their symptoms. I think because they knows that this datat is valuable for the public’s benefit esp. disease prevention, tracking and also disease control and might have a shared benefit on disease screening testing.
However as you said in the presentation. Peoples thought have changed overtime. Nowadays, most of Thai people think that we are not have covid in Thailand. I saw some patients dont want to share their health data anymore (or they might be fatigue to response to questionaire). They dont share their URI symptoms on screening paper, they not share their travelling data by scanning thai chana application.However, in my opinion, I think the perception of people in data sharing is not change. Because there are still not have an event of privacy data leakage. They are still welcome to share their health data that have public’s benefit,but less sharing might be from fatigability and less concern.
-
@Thanachol, your dashboard is showing the minimalism style; simple but most informative. My first impression is your dashboard is very tidy. It is easy to get the important data and also, I love the highlighting on the matrix table. This brings the country of interest to stand out while the others blending smoothly with the background.
However, I think we face the same problem that our dashboard can not automatically update. I try to use the technic as arjarn shown in the lecture but it didn’t work. Can anyone help?-
I go back to the Power BI desktop and manually update the database. The dashboard in the desktop app is updated to 29/6 as expected. So I think the problem is the published dashboard can not update their database automatically.
However, on Than soo soo dashboard, his published dashboard can automatically update to 29/6. please share me how can you do that.
-
-
Please visit my dashboard at: https://bit.ly/2YJzSPw
In my dashboard, I try to summarize each category of cases such as confirmed, death, and recovered cases. I use the same color for each category of cases in order to easily recognized.
The main idea for the first page is the global situation. The map in the middle of the dashboard is to locate the country of interest. The table on the right-hand side is helpful for viewer to navigate their interest and they can sort countries of their interest by number of confirmed/ recovered/ death and also the country name.
On the next page, I try to summarize the continental situation and the movement of the confirmed/recover/death cases across the continents by 100% stacked column graph. -
Please check and comment on my dashboard: https://bit.ly/2NQKgip
The map is a powerful visualization for interested people to easily understand the overall global situation. In combination with the card visualization, they can summarize all useful data of COVID-19.
The funnel chart and treemap chart are helping in tracking progress and showing in hierarchical data.
However, I think in terms of tracking progress the cumulative scatter plot with play axis can show the interactive data and easiest for understanding. -
Please visit my dashboard via this link:
On the first page, I would like to show the overall situations all over the world, and the next page is to show the situation classified by continents for making more clearer how COVID-19 disease spread across continents.
Then the third page goes deep down to details of the top 3 affected countries. I think people want to know when it will end, especially for the most affected countries. The fourth page shows the sparkling line of the top 10 countries with confirmed cases, people can see the trend that some countries have good control of disease and there is a low number of new cases while other countries are not.
The last page shows the dynamics of daily recovery across continents on a monthly basis, the cumulative of death case in each continent, and also predicted model of the global situation in the next 30 days. -
Hi everybody,
Please visit my power BI table and matrix at
https://app.powerbi.com/view?r=eyJrIjoiZGM5NmIxNzEtMTFhOS00YzdmLTgzMzAtOThmYjMwZjk0ZGZlIiwidCI6IjliYzU4NWY5LWE4YjgtNDMxYy05MDEzLWVmYTdiMmI0MGNkZiIsImMiOjEwfQ%3D%3D
Most parts of the table and matrix are the same as others. The differences are, I decided to use the maximum number of confirmed, death and recovered cases. As the provided data source is a cumulation of cases on the most updated day, so we don’t want to know the sum of the cumulative but the user wants to know the cumulation of cases on the specific time period. Another thing is I put the GDP under each country in the matrix table because GDP does not change overtime period. However, I agree with Thanachol that GDP should put in a separate place because it does not affect much on audience perception and decision making. -
I choose this dashboard.
https://covid19.who.int/
I like it because it has all data under my interest, and it very updates data (as I access at 10.58 pm, the website shows the last update at 10.55 pm).
First of all, it contains a lot of information from globally, regional to the country level.
It includes the report of the daily new case, daily death case, and also the cumulative number of new case/ death cases.
Secondly, I like the function that, viewers can stay in the same graph but you can change the variable of your interest, example for, viewers can select the variable “cumulative new cases”, or “daily new cases.” Even the graph wants to show the trend, I think the histogram is also easy to understand.
Case comparison by region, I love the graphic that the viewer can highlight the region of their interest. However, instead of the stacking histogram graph, it can be shown in the line plot to see the trend of cumulative cases in each region.
For the situation by country. I like it because viewers can change the variables of the graph, it has completed data for one country in one graph, also the exact number of each day by putting the cursor on the interested date. Nevertheless, I think it should be better to identify what is the see-saw line or the curve line means. The graph that selected to present here, on the first page, I think they selected from the most cumulative confirmed case, however, I think it should include the most active country in every region.
Finally, they use a constant color for each meaning; the blue color represents a new case, orang color represents a death case. For the regional comparison, they use the same color to represent each specific region. -
Thank you for your question, it is very interesting and challenging.
In this study, the author studied epidemiological data of dengue cases in Honduras. If I were a researcher, firstly, I will clarify the case definition of dengue. There is an overlapped presentation of 3 arboviral diseases and it is difficult to do a serology case in all suspected cases, however, the clearer case definition will help the reader to determine the reliability of data.
Secondly, the incidence of dengue cases in the study area is not enough for making a public health strategy to prevent further spread of disease. I think I will integrate other factors into this study, as you mentioned, the rainfall and also the density of population, habitat and, a local public health center that might affect the dengue health literacy of the population.
For severe dengue cases, I want to add the GIS map of their location and number of hospitals. Because we don’t know what cause a lot of severe case in this outbreak, it might be from the host, dengue virus strain itself or, from difficulty in access to health care services -
Nice question Thanachol, in this study the author uses the numbers dengue cases from the National surveillance system and they didn’t mention anything about the case definition, the only thing that they gave us is the National surveillance system define the dengue cases by syndromic and laboratory surveillance.
For the second question, the most common 3 arbovirus infection has overlapped in clinical presentation. In cases with high-grade fever without any change in the laboratory results can be dengue fever, Chikungunya, or Zika virus infection. Furthermore, all of these 3 viruses also have a co-infection. So, the clear definition of cases is very important and it impacts the prevalence of dengue cases.
In the previous study by the same study group, when they use a laboratory result to define the case. They reported the contrast in the number of cases of dengue and Chikungunya in the same area. In an area with a high prevalence of dengue, the prevalence of chikungunya is relatively low. In contrast which area have high prevalence of chikungunya, it will have a low incident of dengue. -
Hi Penpitcha, as you mentioned in the very first of your presentation that clinically significant drug-resistant TB is classified into MDR-TB (resistant to isoniazid and rifampicin) and XDR-TB (MDR-TB + resistant to fluoroquinolones and one of 2nd line drug).
In your presentation, you didn’t mention about incident and location of rifampicin-resistant TB. Is your study have data about this kind of TB strain? If your study not mentioned about this, do you have an explanation why don’t they report this? I think this data might be very important to identify patients with MDR- or XDR-TB and it might have an impact for public health workers to set the strategy to solve this problem. -
Thank you Arjarn Chawarat, I understood that student who submitted the late assignments will have a penalty. However, for peer who didn’t receive the presentation on time or even the presentations who did’t receive questions on time, will you let a 24 or 48 hours responses after submission as an on time response?
-
This study emphasizing fall incidence that happened during the study period mapping together with service location. The study told us that the incidence is increasing over the time (that might be from more EMS calls). As we all know, the ultimate goal for falling problem is fall prevention. The incidence is just the beginning step of the fall prevention strategy. However, I have one curiosity, If you were the researcher in this study, and you can redesign the study again. what data will you add to this study to give the audience more information to create a strategy for fall prevention?
-
-
-
Chanapong, I like the way you find the research article, both of them are interesting and their research question is amazing! Firstly I love the second article; the nighttime light and prevalence of hypertension. However, when I look deeply into the content I preferred the first article about falling rather than the second one. Thailand is going to be an aging society, falling is the most important cause of morbidity for the elderly. This research matched the data of fall prevention education, an incident of falling ( represent by EMScall for falling), and elderly population. At the end of this study brings up a lot of recommendation for mapping method and how to present the data effectively.
-
I think the first one is better than the second one. Even the study is quite outdated, but TB is one of the most important health issues in Thailand. This article mentioned epidemiological data of TB and the authors can selected the data to draw a figure interestingly. For the GIS picture can present a lot of words, the treatment success rate contrasts with the newly notified case. This brings up the new question “What strategy that each province implement for getting rid of TB, which one is the best that the policymakers have to choose as a model for other provincial?”
-
Great job Penpitcha! I like both of your selected papers. Both of them open my eyes to the use of GIS in combination with epidemiological data that brings a new research question. Both of them are interesting. However, I prefer the second one, the same as Pyae Phyo Aung. Tuberculosis is a big health challenge for Thailand. This research is timely, the way of data presentation is interesting. The GIS pictures bring up a new research question such as “what happened with the TB treatment in this area, why the number of patients with MDR TB increased, and the prevalence is scattered to other areas?”
-
-
-
-
-
-
-
-
-
This CRF is good. I have a bit more suggestion to make this CRF perfect!
1. The date and time format
I think it will be better if the CRF has a suggestion for use B.C or A.D because the four digits for the year can be B.C or A.C and for the time format it will be better to use the 24 hrs. format
2. The ID number
This study has only 2 sites. I think it is better to have only 1 or 2 square to fill for study site ID. In addition, if I was a data entry person, don’t understand what is the student ID stand for?
3. Some abbreviations are not standard use such as “unk”. It might lead to miscommunication and false data entry.
4. The height is in centimeters, I think all of us have at least 100 cm. So the square in the height column should have 3 squares
5. I think the CRF should have eligibility criteria check section after all of the eligibility criteria were check ( I mean that the UPT and physical exam should come before eligibility criteria check section).In addition, the physical examination, I think it would be better to list on the same page in order to make this CRF more user friendly.
6. Vaccination arm should be 3-valents or 4-valents because in this study we didn’t compare with placebo. And the Ab titer should have a space to fill the value of the titer results
7. If we didn’t focus on the race other than Asian or non-Asian, I think it doesn’t important to classified the race to 6 category -
-
-
-
I think the main difference is the personal adaptive mechanism. For me, initially, I also felt struggles with the new system because I have to type the patient’s hospital number for every single patient, and the system always down when a lot of physicians start to work at OPD and retrieve patients’ data at the same time.
After I got familiar with the new system and I perceived that the IT unit put a lot of afford into it and provide a prompt, easy access consultation. I opened my mind to notice the good thing about the new system and I found that the use of a semi-EMR system does not disturb human contact with my patient. -
-
-
In comparision with the up-to-date data on 30th June, the predicted confirmed case is lower than the real situation. However, the death cases and recovered cases are in the predicted range. I think this comes from the rapid spreading in the US (which is take a very long period of disease spreading and still not yet reached the maximum point), South America, and Russia.
-
-
-