Forum Replies Created
-
Posts
-
-
Hi Professor, since the objective of my modeling is just to update the paper with the transmission projection, I think the introduction of an intervention/s (eg, PrEP, counseling, barrier methods) would be outside the scope of the project.
However, for the purposes of the discussion, I am considering PrEP as the primary intervention to reduce HIV transmission among genders. It is easier to estimate PrEP usage since it is dispensed on HIV clinics and is controlled by the government. There are also trials indicating the efficacy of these prophylaxis therapy. On an SI model, this intervention interrupts the conversion of a susceptible individual to being infected – assuming that all susceptible individuals adhere to the therapy. Challenges persists on medication adherence such as supply constraints, stigma of going to hygiene clinics, stigma of taking medications for HIV prevention.
Reference: Consolidated HIV guidelines for prevention, treatment, service delivery & monitoring
https://www.who.int/publications/i/item/9789240031593 -
In the paper of David & Gomez (2014), they used the SI model for HIV. I think it is rational to use the SI model since the states of HIV can be susceptible and infected. Although there are accounts of about patients recovering from HIV, these cases are very rare – in fact, these remain a huge mystery for scientists. In addition, although there are HIV medications that can suppress the viral load to undetectable levels and proclaim one to be “HIV-undetectable” who cannot transmit the disease, incorporating a “recovered” state would be erroneous as the patient is not completely cured from the disease.
Variables:
1. Beta-1, Beta-2, Beta-3: Transmission rates for each gender (heterosexual, bisexual, homosexual), respectively. Data: Infections rates estimated by David & Gomez (2014) (from 1990-2012) and I am continuing search for updated values or if there are published values in literature.2. Infected (I): Initial number of infected people
Data: 56982 people with HIV registered taking anti-retroviral therapy
Source: https://doh.gov.ph/sites/default/files/statistics/EB_HARP_January_AIDSreg2022.pdf3. Susceptible (S)***: Number of susceptible people
Data: 72.02 million
Sources:
https://data.worldbank.org/indicator/SP.POP.1564.TO.ZS?locations=PH
https://popcom.gov.ph/popcom-number-of-filipinos-in-2021-estimated-at-110-8-million-sizes-of-families-trending-lower-at-4-members/#:~:text=POPCOM%3A%20Number%20of%20Filipinos%20in,Commission%20on%20Population%20and%20Development
https://doh.gov.ph/sites/default/files/statistics/EB_HARP_January_AIDSreg2022.pdf***Of the 110.8 million Filipino people (source: Philippine POPCOM), 65% are in the age of 15-64 years old (source: WorldBank). It is assumed that this population group are comprised of people actively having sex. David & Gomez (2014) also looked at this specific population group.
-
Hi Professor, I’d like to do n sexuality-based predictive modelling on the HIV epidemic in the Philippines. There has been an initial work (https://philsciletters.net/2014/PSL%202014-vol07-no01-p062-066%20David%20Gomez.pdf) which was published in 2014. I’d like to input newer data or tweak the model based on the latest set of data from the Philippine HIV registry. The probable research goal is to provide supplementary evidence on the key population whom to provide the interventions to mitigate the spread of HIV.
Do you think this is feasible within the bounds of the lectures provided on this course?
-
I can perform risk assessment first to determine if there is a need to transition to cloud computing or retain for a couple of years.
As mentioned in the lecture, transitioning to cloud computing can be costly upfront. Can the resources of the hospital afford this transition? Will the users be amenable to this change? Is there a need to transition to cloud computing if the current system is working well? How will this transition affect the current operations? Are there vendors in the market who are reputable in handling health data? These are some of the questions that need to be answered first before considering the adoption of cloud computing in the hospital system.
-
Based on experience, these pop-ups will redirect to some phishing websites (appear legitimate but if you look at URL it’s not legitimate, for example, a banking website) download an infected file, or infect the browser by installing numerous add-ons which will cause the browser to crash. Sometimes, these add-ons collect input data from users which may contain sensitive data.
Also, I haven’t heard or seen a browser-based scanning of computer for virus and be able to clean it if there is. It’s most likely a deceptive and malicious advertisement and will not click on it for security reasons.
-
In the workplace setting, there are several logical and physical security controls in place to ensure security of the system. For example, company issued laptops are forbidden to install any other applications not approved by the company (an administrator prompt will pop-up). Other logical controls include password refresh every 90 days, use of alphanumeric+ symbols. In terms of physical security, our office is equipped with access cards such that only employees can enter the room. In terms of online security, intensive spam filters are set and any external emails are automatically tagged as “external” with warning messages. So far, I have not encountered any security breaches in the workplace.
Meanwhile, I had encountered several instances where my PC got infected with virus and worms. It was rampant during the early days of the internet, like the use of LimeWire to download songs (files disguised as songs), downloading .exe files on some website offering a cracked version of a paid software. Flash drives were also the common source of infection, particularly from public school computers and computer shops. Sometimes, antivirus works however, I had experienced wherein I had to completely perform a hard reset (reinstall Windows) on the computer because it was too slow.
As preventive measures, only download software from legitimate developers. Enable auto-updating of definitions of antivirus programs, perform quick scans on your PC from time to time, visit websites that look legitimate and inspect the URL if you are really on the website you are supposed to visit.
-
In the context of COVID-19, dashboards became helpful in decision-making among policy-makers. It quickly displays the daily reported confirmed cases, deaths, ICU/ hospital utilization among others. We can also implement an in-depth look on these data by implementing filters to determine which areas are of high, moderate, or low risk such that allocation of resources can be efficiently distributed, and appropriate public health measures can be implemented.
Some of my insights when designing a dashboard are first, it must be intuitive for the user. Overloading with information without context makes the dashboard practically useless. Second is to display with the relevant information needed. For example, a COVID-19 dashboard may not need the capital of each country or GDP when you only want to see the daily reported numbers. Third is the choice of colors and fonts. The fonts should be readable and colors should suggest on what the data are suggesting. For example, a decrease in GDP may be indicated with a red bar or red font to emphasize something. Fourth is the ability of the user to use filters based on their needs. When working with big data, a user might only want to check for a specific subgroup.
-
If I were to build a mobile app, I’d like to develop a game with in-app purchases. Games are always on demand from gamers; however, you should be able to bring something new to the market. In-app purchases will enable return of revenue on time and resources spent on developing the app. On gaming, it is preferred to have a native app installed on the phone to fully maximize the graphics and resources of the mobile phone. Although there is an innovative approach where you can “stream” games from the cloud, issues on connectivity highly impacts the gaming experience.
-
If I were to build a website, I think I would choose to re-learn HTML. I had taken an elective way back in high school on creating a website using HTML codes. I remember it was an overall fun experience. With the new iteration of HTML, namely HTML5, for sure there will be lots of things to learn again.
Also, one field that interests me is Data Science. I think Python is one of the most commonly known programming language in this field and would benefit me should I consider my interests of pursuing that field as well.
-
Information technology enables relatively quicker access to healthcare. In just a few tap, you can have a doctor’s appointment booked, consultation through online platforms, deliver medicines, receive laboratory results in few minutes, and much more. In the old days, people have to go through face-to-face visits to avail these services but with the power of IT, it has saved so much time (which is intangible) for patients as well as practitioners. Through IT, it gave more patients the medical attention they need more than ever.
In the Philippines, access to the internet still remains an issue. In some areas, there are no reception at all, while some are still stuck in 2G/3G connectivity which significantly affects the transmission of data. It also raises the issue of equity since not all people are equipped with smart devices to connect to the internet due to financial and/or literacy concerns. Even in urban cities, there are still areas where there are “dead-spots”. Also an issue would be how data are being handled by the ISPs – what gets shared and who gets and receives the data.
-
There are four network types based on geographical area. In general, network type varies on the distance the network is intended to be used. In general, the cost increases as we cover more distance of interconnecting devices.
1. Personal area network (PAN) are those that can be implemented on the personal level like use of Bluetooth and wireless connectivity. Smartwatches can be a great tool if we want to monitor blood pressure among participants. The main limiting factor in using this type of network is its very small working distance.
2. Local area network (LAN) are those that are implemented in universities, hospitals, corporate setting. These can be used in connecting systems on an institution. Cost of interconnecting different departments in an institution maybe costly but I think it will be worth it on a long-term basis.
3. Metropolitan area network (MAN) are like LAN but implemented on a city basis. Usually internet service providers (ISPs) are responsible for interconnecting different parts of the city. People can be controlled who can access which and what data they are accessing.
4. Wide Area Network (WAN) – a great example of WAN is the internet. But in recent times, internet has been a source of questionable data and hacks from various sources so security may be an issue.
With regard to the question, I agree with some folks regarding the distance of the data source to MOPH. I think it would be best to use PAN (for transmission of data from wearables to a computer system, like mobile phone) and MAN (for transmission to MOPH). This also considers security of data transmission.
-
In my country, a lot of government agencies need to adapt and reshape their organization for the digital future. Previously, filing taxes all resorted to manual computations. It was only in recent times that there were attempts to digitize filing of taxes (through installation of software and electronic submissions) but there are still prevalent problems like unsuccessful electronic submission due to unknown error, and most importantly, there system is there but there were no guides on how to fill-up the form.
Another example is our COVID-19 contact tracing application. In the earlier days of the pandemic where contact tracing methods have sprung up, individual cities implemented their own contact tracing applications. This means that whenever we visit other cities, we have to download another application to accomplish the task. It was only after a year into the pandemic where the government mandated the use of a national contact tracing application. However, some businesses were wary of its use and resorted to their own respective applications.
-
Hi all,
This is my dashboard for the project.
Kindly look at the first two pages only.
I opted to deviate from the JHU dashboard in terms of the color scheme since it was to dark for me. Provided data limitations (ie, vaccination), I opted to simplify the dashboard with only the necessary visualizations included.
This dashboard contains the necessary details to give a user an overview of the COVID-19 situation, that can be filtered by date, continent, and/or country. Map was used to illustrate spread of COVID-19 across the world. Separate bar graphs were also constructed to illustrate confirmed cases, deaths, and recovery. These three parameters are standard across all visualizations for COVID-19. On the second page, I put a 7-day moving average measure since it is one of the epidemiological measures that can be used whether the disease is spreading or contained. Perception-wise, I chose blue for confirmed cases to give some calmness to the user even though the cases are skyrocketing. Deaths were symbolized by red which is usually associated with danger. Recovered was illustrated using green color to mean “go” on with life.
Hope you all like my dashboard and I appreciate your feedback!
-
Hi everyone,
Here is my dashboard for this week’s assignment.
Allow me to explain the graphs and illustrations which are separated by tabs (kindly skip to Page 2).
Page 2
These graphs illustrate the bar chart, map graph, and multi-row card visualizations, respectively. The bar chart us group by continent. The map graph visualizes the “spread” of the confirmed cases across continents. Lastly, the multi-row card provides the instant look for each country case count. Overall, these three charts/graphs provide an overview of the cumulative count of COVID-19 cases worldwide.Page 3
The graph illustrates the ratio of daily deaths and confirmed cases of COVID-19 per country. This will give an idea on how many deaths were reported per day, in relation to the number of cases detected in that day. Usually, peaks in the graph indicate a surge. Since COVID-19 case reporting a sometimes delayed, the effects of death can only be observed a few days or weeks after the surge has passed. This can be an indicator of how deadly the surge was.Page 4
Funnel plots are usually best when there are two variables being compared in a same measure. For example, for an age group 10-20, it would be good to see the ratio of males:females in terms of mortality rate. However, the source data has only limited information. Hence, for this visualization, the GDP grouped by continent was presented. Same goes for the pie chart since there are no other parameters that can be presented using a pie chart.Page 5
The line chart shows the cumulative confirmed COVID-19 cases as compared to deaths through time, categorized by country. This is a standard visualization across known dashboards for COVID-19 as it provides a good overview of how fast a country tallies confirmed cases and deaths through time. An added feature, forecasting, was used to project the course in terms of case counts and tallies for a period of 6 months. This will help in preparing the amount of financial and healthcare resources in the coming months.Page 6
Blank page, ignore.Page 7
This line and clustered column chart visualizes the 7-day moving average by country. 7-day moving average is a usual indicator in epidemiologic trends particularly on diseases that have incubation time such as COVID-19. Also, 7-day moving average somehow balances the trend by protecting it from random influxes, eg. testing laboratories are closed on Sunday. This was calculated using “Measure” feature of PowerBI.Page 8
Sparklines indicate trends across time. For this visualization, I used a Matrix with continent as a main grouping variable, followed by countries. Sparklines within matrices provide quick comparison across region (or continents) in terms of confirmed, deaths, or recovered COVID-19 case.Page 9
Treemap was used to visualize confirmed COVID-19 case of a country as a ratio of the whole confirmed case counts. This provides an overview of the COVID-19 situation as well.Page 10
A good comparison for a scatter plot would be the plot of total number of tests conducted and total confirmed cases. However, due to data limitation, deaths vs confirmed cases was plotted instead. This plot gives an idea on the spread of datapoints across axis – whether the give off the same ratio as other countries or as an extreme outlier. This graph could be perceived as an extension of stacked column chart where it can give information relative to other countries. -
Hi everyone,
Attaching my PowerBI dashboard for this week’s assignment.
COVID-19 dashboard with number of cases (confirmed, recovered, deaths) is one of the measures in determining the country’s response to this public health emergency.
Best,
Arwin-
For some reason, even after I followed the steps on publishing the report, it still displays incomplete columns. But when in editor mode, I can still see the columns for the deaths, recoveries, and confirmed case.
See attached screenshot.
-
-
This is the Philippine Department of Health (DOH) COVID-19 dashboard.
https://doh.gov.ph/covid19tracker
In brief, it provides the daily updates on COVID-19: case counts and visualization like graphs and line charts or combination of both. The first dashboard demonstrates the data based on case count, the second provides data on number of tests conducted and positivity rate, and the last one summarizes the healthcare utilization indices.
Overall, the color palette follows the DOH brand book, which is a style guide for infographics published by the department. This streamlines the look and visual appeal of the dashboard to establish association. Its biggest strength is its interactive interface. It supports display of data when pointer is hovered to a specific timepoint in the graph.
Colors were generally light and can easily be perceived. It follows the contrast rule of visualization on the green color. The red color signifies attention to rise in positivity rate over time.
Line and bar (simple or stacked) graphs were used to demonstrate trends which is very useful in considering the epidemiologic situation of a particular place, region, or as a whole.
A data drop is also available for download in case further visualization is warranted by the user.
I guess one point for improvement is to clearly label the three dashboards instead of drawing context based on the content.
-
Hi Napisa,
Overall, this CRF is nicely and comprehensively done! General comments are concerned with directness of the questions/instructions.
I am attaching the annotated PDF for you to see my comments on different portions of the file. Let me know if you cannot access the file. 🙂
Alternatively, you can access the annotated PDF through this link: https://1drv.ms/b/s!Alz_4z9rkOOTk-5M7q67v8RaurPmmw?e=VmKq9a
Thanks!
-
Some fields have units (eg, pulse rate, temperature), while some don’t have (eg, blood pressure, weight). My colleagues have covered most of my observations as well.
Perhaps, in Physical Examination, it’s too open ended if an investigator tries to specify the abnormality. Perhaps, choices based on the organ system would be a great start in classifying abnormalities observed during physical examination.
-
-
During our undergraduate program, we have a activity called Computer Validation. The main objective of the activity is to digitize input data from printed manufacturing records after tablet production.
We were free to use any software, but I opted to use Excel, with expanded features through VBA (for user specific sign-in and audit trail). I was the lead on constructing the program, but I have sparse experience on coding so it was a steep learning curve.
The Quality Assurance Lead assigned credentials to every member of the production team to prevent unauthorized sign-in. Specific roles have different views (eg, those assigned on the early stages of tablet manufacturing can only view the sheets during early stages). During data inputs, there were Data Validation features in Excel which were helpful during such. For example, when the field requires numbers, we limit the field inputs to numbers only and will get an error message once letters were inputted.
During audit trail, we coded using VBA to register any form of changes in the Excel sheet associated with the logged-in account.
The program was initially tested only on a single computer but we managed to make it work through OneDrive to make it accessible on other computers, of course, after authorization from the Quality Assurance Lead.
We were only given a week to prepare the program. Admittedly, there were bugs in the system but provided the short timeline, I think we did an impressive job in building this software.
-
Remote delivery of healthcare is inevitable, provided the advancements of technology. Its prime advantage is accessibility to healthcare – a promise brought about by technology advancement. In this time of COVID-19 pandemic, telemedicine enabled access to healthcare without physical presence on hospitals – preventing possible transmission of the deadly virus.
However, it is also faced with concerns primarily focused on data privacy and security, as well as inability to assess the true status of patients in terms vital signs and nonverbal gestures. These are important in arriving at a more informed diagnosis.
-
I do not have experience on being an associate for clinical trial. The nearest experience I had was on protocol development for a systematic review on a certain cancer drug. Basing on the steps, it touches mainly on the “Protocol Discussion” part were it was a collaborative process among stakeholders (eg, experts from various medical societies, pharmaceutical industry) as we aim to capture the needed variables and outcomes when we start our assessment.
-
When I was still in the our health department, I was assigned to investigate the effectiveness of high flow nasal cannula for COVID-19. The primary purpose of the review is to see whether if it is worth funding by the national health insurance agency considering the uncertainties brought by COVID-19.
One of the sections of the review is to gather around the experience of front liners who had first hand experience on the technology – meaning, primary data collection was required. We prepared a standard set of questions and an informed consent on the purpose of the interview, how their answers will be used in the generation of the report. We invited experts from different societies and hospitals and gathered information through focus group discussion.
After the data collection, here were some of the problems we encountered:
1. Difficulty in transcribing some medical jargons used by doctors and respiratory therapists.
2. Some experts have contradictory opinions and became difficult to reconcile during writing.
3. As we also sent a questionnaire, some experts who weren’t able to attend the focus group discussion opted to send a filled, written questionnaire. They weren’t able to answer some parameters; hence, there were missing data inputs. -
Hi, sharing my week 3 assignment.
Alternatively, you may visit this link: https://prnt.sc/z4yru23bv3mx
-
Please find my summary on this link
-
Hi everyone, here’s my wrap-up for this week. 🙂
(if the image is not embedded, click this link: https://prnt.sc/7jv2PW_T6hyW )
-
At this point in time, I can say I am a frequentist since those are the methods I was taught in college. Hence, I was conditioned to be a frequentist. Perhaps, this is to standardize “answers” since Bayesian depends on your opinion – to which, opinion may vary from one person to another.
But after discussions on Bayesian, I think there will be instances that one method is “better” than the other. I might be wrong on this position but this is how I currently see this debacle.
I agree with Auswin that exploring each approach will help us understand the differences.
-
Hi everyone,
My name is Arwin Onda. I currently reside in the Philippines and is a pharmacist by profession. My most intensive statistical analysis experience was during undergraduate thesis where we used mathematical modeling to predict the best formulation of a drug. I currently work as a health researcher/ medical writer. I read tons of journal articles everyday and appraise them based on methodology and results. I have basic knowledge on statistics but would like to reinforce (and learn more) that knowledge through this course.
-
-
You can ask them through interviews to ask the behavioral, cultural, or environmental factors that affect their use of bednets. If you are a more sociable researcher and easily get along with foreign people, participant observation may also be useful.
Quantitative research gives you the “how many”, and qualitative research gives you “why”.
-
-
-
-
I can’t remember how many times I’ve lost my files because I need to reinstall Windows! It was such a pain to set-up everything from scratch. Backups weren’t a thing before, good thing external drives are getting cheaper nowadays! I agree on using genuinely licensed programs to prevent any sort of malware infecting the system.
-
COVID-19 dashboard, especially from JHU, has been one of my favorites when I was still working with our Ministry of Health. In enabled me to get data across various countries with ease. I definitely agree on your aforementioned considerations when building a dashboard. Context and appearance are essential in drawing users to actually check your dashboard.
-
-
-
-
-
In Philippines, I don’t foresee centralization of databases (pharmacy, hospital) in the near future. Most establishments are privatized and would cost a huge investment if they transition into a central database. Some renowned hospitals may be able to adapt, but small players may have difficulty in implementing such.
-
-
Hi Napisa,
On the second page, I think there will be mislabeling on the graphs if you use the slicer to filter by country. I like the idea of using a slicer on case counts but I think it would be more helpful if the numbers were categorized (eg, 1000, 10000, 1M) than using a slider.
Overall, nice work! I also like a simple looking interface.
-
-
Hi Professor,
I think it took a good few minutes for the dashboard to refresh. It now displays the updated columns based on my last update.
Still attaching the PBIX for your reference.
Best,
Arwin -
-
-
-
I was considering to be a CRA at some point. I deem it as challenging because you will have to perform side-by-side check on protocol vs what has been or had been done – if they were following the protocol or deviations. I do enjoy performing quality check, but it must have been a headache when there are deviations!
-
-
