Forum Replies Created
-
Posts
-
-
-
-
Dear Phyo, I believe that your case report from (CRF) is comprehensive and thorough and well-designed. I applauded your effort and dedication on the assignment. I really like your design, and the way you highlighted the header with a color in each section. Additionally, your instructions are thorough in most areas.
My suggestions are as follows:
1. For the pregnancy test in the exclusion criteria, would it be better if the result of the test should be filled like pregnant or not? Simple ‘yes or no’ would mean two things: test done and result.
2. Is it useful to collect nationality instead of race? Since different races can be a part of a nationality? And Asian is not a nationality as well. The box in nationality and ethnicity is quite confusing in the form.
3. Height in cm. Since average human height are three digit like 130-180 cm, I believe providing only two box for that area would not be enough.
4. In the medical history, you are collecting how well they control their chronic disease. So, how should a data collector enter the data if the participants do not have the chronic disease? ‘Not’ means not well-controlled or do not have disease? It would be better if those variables are clearly stated as well.
5. You are collecting Weight, Height in both demographics and physical examination section. Do you have any specific reason for that? If not, one area would be enough.
6. It would be better if you could include laterality in the site of administration like right and left.
7. Including consent sign date and date and time of collection of blood sample would provide more relevant information in your case report form as well. -
-
-
Having data standards can be advantageous in clinical research. Among them, I believe that having clinical data standards can be a great benefit for regulatory bodies to efficiently track and monitor their research and conducts during the study process as it can enhance the regulatory bodies to effectively review the clinical research.
For example, CDISC standards such as SDTM (data structure and tabulation) and ADaM (data analysis) are required standards for submitting the work of clinical research to regulatory bodies like FDA. By using those data standards, the regulatory bodies do not have to spend extra time to decode the data again and reduces back-and-forth questions. And it can also lead to faster review process and faster approval, streamlining the review process.
Additionally, regulatory bodies can use those standardized data for automated data validation to check for inconsistencies, ensuring that the process has less errors. Last but not least, using standards can improve data transparency, as it is easier to audit.
Therefore, I believe that having data standards will benefit both the researchers, regulatory bodies, and the patients by enhancing consistency and transparency.
-
I have only a few experiences regarding the study projects related to data management, but I have been involved in some internal data management projects, not for study purpose.
The projects that I have involved only includes “edit check and some logical check” along with “data backup”. Those projects include both offline and online data management tools or systems. They include logical checks such as having range check so that invalid values cannot be filled, and consistency check like setting primary composite keys so that duplicate variables cannot be entered. In the case of data backup, they copied and stored those copied files in their external hard drive. However, there is no concrete data recovery plan.
Additionally, when the project includes “online” data management system, they include audit trail and time stamp. However, most “offline” tools in the projects that I have involved lack those features. User authentication is a loose process and the access control level is flexible in most of these projects. It contradicts with good data practice.
And the computer softwares that they use to process data includes MySQL server, Microsoft Access, Excel, Google Sheets, and so on.
-
I have learnt responsible research innovation and the importance of taking precaution in applying AI tools in health sectors. Since we are dealing with human lives and life-to-death decision, using AI, a tool that most of us do not understand their inner intricate reasoning, should be considered with great care and responsibility. Besides, the innovation should be centered on human, it would assist us, not replace us in the most important decision-making process.Here is the link to my infographics. https://snipboard.io/5IGiQe.jpg
-
My recent data collection experience was in 2024, it was for my public health related research. I usually follow similar steps of the data management workflow but there are some important areas that I did not do in my previous work.
When initiation my data collection, I wrote down what steps to follow and what inclusion criteria should be considered along with which type of variables are of interest to my study (Protocol Discussion & Data Design). Then, I determined that the data collection would be via electronic data capture (Data Acquisition).
Before doing any data collection, I wrote down how the data will be stored and managed after data collection along with the plan to archive it. Instead of developing CRF, I developed structured questions which are open-ended and relevant for my exploratory study. Additionally, I did not collect any identifiable variables like their emails and name etc.
I created the Google form for the data collection and made sure that only I can see the participants’ responses (Database Access Control & Database Setup). The thing that I could not implement in the database step-up is installing proper checks on range and consistency checks in some questions since they are open-ended and the limitation of the usage of google form. But I was able to ensure there were presence checks for the completeness of some important variables (Edit Checks).
Then, I checked my google form with a dummy entry before rolling it out (Data Entry Test).
I could not implement programmed data validation checks, but I manually validated the collected data. Since the data was captured electronically, I have some advantages like not having illegible text (Data Validation). But since it was open-ended questions, and we did not collect any identifiable variable, it was hard to control the quality of data like in the case of quantitative data (Data Quality).For Data Quality Assurance, I did not put much effort in my previous work and there was no external data for merging as well. Additionally, I did not have a proper written analysis plan before any analysis. I did plan the analysis but did not formally write it down step by step.
I have archived my data from google form and locked it, only after the data analysis and final report have been submitted (Data Manipulation and Analysis, Study Report, Database Lock, Archive).
To conclude, if I would re-vise the study, I would put more efforts and attention on data checks, quality control, and quality assurance. I would implement proper range, and consistency checks in designing the database. And I would outline the written steps of my analysis before performing any analysis.
-
I have collected numerous types of data for different purposes during these years. It includes both qualitative and quantitative data for purposes like research, exploratory studies, and other for work related purposes such as feedback form.
In 2024, I collected qualitative data on selected participants in the border area to explore their access to health care in relation to their difficulties. So, the purpose of the data collection was for exploratory research.
I conducted data collection through semi-structured surveys via online and simple data collection tool, Google form. And the data collected was specifically for the research. So, it was primary data collection.
Additionally, the method of data collection was self-administered. Even though I generally explain the content of the questions in the survey, they had to fill their form in their free time. So, it can be regarded as self-administered.
The problems that I encountered included having low-response rate, and having selection bias due to the nature of the research. Additionally, since there were no advanced data check and validation process installed, there are some errors in the data such as wrong answers to the wrong questions. Though all questions are made as required, some participants only filled the minimal data in such fields. So, those questions have incomplete answers. Since I circulated the survey form and requested each participant to share it with their friends and relatives, the research design has selection bias.
-
When developing a disaster recovery plan for my organization with varying technical skills, limited budget but medium-scale critical operations, it is important to balance between cost-effectiveness and disruption of services.
Risk Analysis and Assessment: It is the first thing I should do to analyze and identify the external risks such as disasters, facility risks such as power outage, data system risks such as data backup failure, departmental risks, and desk-level risks. Then, calculating the likelihood, impact, and restoration time along with associated risks can assist me in prioritizing and preparing necessary procedures in my recovery procedures.
Anticipated Effects: The second procedure is the identifying the anticipated effects of the disasters along with the cost of downtime and lists of anticipated affected entities. It is the process of matching potential risks and anticipated effects on the system.
Evaluation of Disaster Recovery Procedures: After analyzing the risks and anticipated effects, it is time to evaluate the existing disaster recovery procedures and mechanism such as spare server, multiple power supply, and backup of critical data. When evaluating them, it is necessary to balance between cost and impact of those systems.
Organizational Considerations: Since the organization has limited budget, sometimes, it is not possible to choose multiple sources of power supply to the whole system, but it must ensure enough power supply for the critical data system in those times of disaster. Instead of backing up the full data, backing only critical data and replicating them can be an effective in the case of downtime while cutting some budget.
Additionally, an organization with the people with varying technical skills can be sometimes challenging to respond timely in case of an emergency. So, it is necessary to have a list of focal point of contacts from each department that are essential in responding a disaster. Besides, providing training on cybersecurity and data security and mock drill can prepare them for a potential disaster.
-
Hello Cing, thank you for sharing your comprehensive infographic. I agree with you that even though telemedicine can bridge the distance and access to healthcare in rural areas, there are some things to improve like digital literacy and divide, the one you mentioned in the “risks”. Proper training and awareness to digital literacy can narrow this divide.
-
Hello Wannisa, thank you for sharing your insightful infographic. I agree with you that providing something free does not grant you the immunity from accountability and legal obligations. Additionally, having a relevant law and regulations might deter some incidents like the one mentioned in case study.
-
Hello Cing, thank you for sharing your insightful infographics. Like you mentioned in PDPA part of your infographics, I think having an act like this can enhance security measures and required protocols in cross-border transfers of data. Additionally, stating penalties is also a good thing I like about PDPA, as it outlines PDPA compliance process.
-
Hello Tanaphum, thank you for sharing your insightful infographics. I like your suggestions on the ways to delete both paper-based and electronic records related to healthcare. Using software-baed deletions and destroying both physical hard drives and encryption keys are the ways to ensure proper deletion. However, I think shredding would take a lot of time if the paper-based records are huge. Alternative ways would be boiling.
-
Hello Kedsarin, thank you for sharing the colorful infographic. I like your suggestions in the lower left corner of your infographic. I understand that telemedicine is still not up to the traditional standards, however, we can ensure its effectiveness by outlining follow-up protocols and more digital literacy training to the care-providers (like you mentioned in your info-graph).
-
Hello Cing, thank you for your detailed suggestion. Yes, I agree with you that weight is not directly associated with the study. For the vaccine, I think adding site and lot number would be beneficial as well, but I believe route is the same for all vaccine in this study (Intra-mascularly). For the sample type, yes, it is quite confusing in my case report form.
-
Hello Aye, I agree with you that having standards and agreed protocols can improve data interoperability in terms of research as well. And mostly in research, it can enhance research regulation and review process. Since there are millions of papers generating each other, it would be easier for the regulator to have a shared standard to make the process more efficient.
-
-
Hello Cing, thank you for sharing your experience and suggestions on electronic tools for data capture. Yes, in the case of clinical and public health data, role-based authentication is a standard choice. However, it is hardly implemented in low-resource settings where the basic digital literacy is below average. So, when conducting a study that relies on the data from those settings, we should be mindful of minimal standards to meet in respective studies so that we do not violate patient’s privacy and so on.
-
Hello Aye, yes, it is indeed a challenging issue when the organization relies on offline data management tools that lacks proper logs and validation methods. However, even though we cannot have an audit trials in excel and other portable office tools, we can at least have some data validation checks in the excel. In that way, we can at least minimize some inconsistent data.
-
Hello Wannisa, I agree with you that both GDPR and PDPA have similar approaches as well as fundamental principles. I like your way of presenting case study, it is very informative as well. However, though I believe that turning to digital storage would minimize such risk of records being used as paper bags, digital storage are also more prone to cyber breach as well.
-
-
Hello Aye, thank you for sharing your colorful infographics. Like you mentioned in the infographics, I totally agree with you that we must have a positive attitude and perspective related to reporting and delivering justice or accountability. Additionally, we must take heed of the nature of work and data we are dealing with and how sensitive and fragile they are so that we must take every possible means not to breach it.
-
Hello Wannisa, thank you for sharing your insightful infographics. I really like your case study presentation in your infographics. It shows the concerns, violations, and actions needed to be taken by DPO. Additionally, I think you reinforced ethical principles with responsibilities. It is a good way to present it as well.
-
-
-
-
Hello Phyo, thank you for sharing your experience. Yes, it is challenging to strictly enforce user authentication methods in low-resource setting and where the technology of the information system is not well-designed. Additionally, most of the data collected was via paper-based documents, and it was difficult to store them securely without easy access by the other staff. By the way, it is good to learn that some NGO are using R software to analyze their data.
-
Hello Cing, thank you for sharing your research project. I totally get the difficulty of hypothesizing the right questions before any analysis. And I believe as well that we should always stick to our first formulated research questions even if our data suggested otherwise. Another alternative is conducting another survey (research) based on the results of the first data, by formulating more suitable research question before any analysis.
-
-
Hello Aye, thank you for sharing your insightful infographics. I totally agree with you that taking precaution before implementing anything is a great ethical consideration especially in sensitive domains like healthcare. Additionally, as you mentioned in the first point, we should consider AI that could empower and assist human.
-
Hello Phyo, thank you for sharing your experience and challenges. I also faced some of the challenges you have mentioned that involvement of several staff in data transcribing leading to data security and confidentiality issues. Additionally, I can agree with you that having data checks like logical and consistency checks could secure more quality data.
-
-
-
Hello Phyo, thank you for sharing your challenges and field experiences. Yes, I agree with you that it is sometimes challenging to record the data in their mother language due to scarcity of proficient translator especially if the ethnic group is a marginalized population. Since it can impact the quality of recorded data, we should minimize it using different solutions like pre-voice recording of questions in different languages.
-
Hello Siriluk, I agree with you that having activation criteria along with a clear communication plan during disaster is an important aspect of effective coordination of disaster recovery plan. Additionally, like you mentioned, we need to consider cost effective backups option as well along with regular testing and training.
-
Hello Aye, yes, it agree with you that assessing and analyzing risks is the first and major step we should take in preparing a disaster recovery plan. Like you mentioned, the committee must be composed of dedicated persons from different departments. Just come to my mind, I think it would also be beneficial for the organization to include public relationship specialist to manage the reporting of the disaster effectively.
-
-
-
-
Hello Aye, thank you for sharing your experience about project management. I can agree with you that allowing frequent non-essential adjustments can sometimes harm the whole project. So, having a clear objectives and clear features can also improve the communication between end users and the developers. Additionally, it is also a good practice to have a review session after a project for better improvements.
-
-
-
-
Hello Cing, I agree with you that problem statement is a important first step in managing a project. Using a gantt chart allows you to monitor and track where you are and what you should be expected at that specific period. However, like you mentioned, when more and more people are involved in the project, it is harder to track using a strict chart. Instead, like suggested in the linkedin course, we could use multiple gantt charts by dividing the projects into smaller chunks.
-
Hello Siriluk, I agree with you that high availability can provide major advantages in times of crisis and outages. However, there is also a need to balance between cost and the type of high availability (HA) we are going to implement. However, in the long run, HA can support hospital growth by improving patient’s satisfaction and system continuity.
-
Hello Cing, I agree with you that we must balance between cost and the level of high availability. However, it is also important not to delay operation of patient’s service delivery. And having a secondary system like paper format can be handy in case of power outage, cyber attack, and internet outage.
-
-
Hello Cing, I think Malaria is a notifiable disease in Myanmar, so I think the hospital can share the data without any identifiable variables with government related surveillance system. However, it will be very hard to do so in the current situation of Myanmar where all the state infrastructures and data flow is severely limited on both sides.
-
