Infectious diseases and clinical pharmacology expert Dr. Benjamin Co has been thankfully breaking down coronavirus numbers in his personal blog since the outbreak started. The perspective he provides is informative, and comforting in those who are craving for a clear picture of how we are faring against the virus. Dr. Co will share daily updates and analysis of the Department of Health reported numbers with ANCX.
Refer to the link DOH.gov.ph or up-to-date data or to COVID19.gov.ph. (The latter is not a secure site.) The new site for the Department of Health is user friendly, provides more information where a COVID19 tracker is seen. Readers can check their official site where Data Drop for raw data can be found.
One useful site is COVID19stats, where one can see most of the DoH data in graph format.
We have had more cases yesterday and today compared to the declining trend last week. The DoH reported overnight 350 new confirmed cases, 89 new recoveries, and 13 new deaths. While there is an increase in new confirmed cases, this can be attributed to the backlogs and the increased testing (but it’s difficult to correlate it statistically because data are latent). Since the erratic numbers come in on a daily basis, I recommend using the seven-day rolling average as it provides us a clearer picture of what is going on and eventually interpreting the data.
It would also be helpful that all the backlog data be separated and evaluated at the end of the crisis in the country. Adding outdated data when the agency receives them and then announces them is confusing to the public. This is also a bad way of reporting latent information. I suggest that a team be formed to separately evaluate all the latent data (i.e., tests, deaths, and recoveries).
How do you determine when to consider them as latent information? If we presume that the clinical course of the disease is 14 days, then we set a time stop from 14 days back. That means all data before May 12 should not be included in today’s reporting and should be considered as part of “backlog” information.
There are two ways of handling latent information. After all the backlog has been collated, they are added to the running total on a singular day (example if they finish collecting these on June 10, then they add it all to the June 10 data). Anything past that date that we included the backlog will correct itself the following day anyway. The other method is to count all the backlog data when we have zero (0) cases reported for seven consecutive days. This way all information provided are real time and reflective of the gains (or losses) of any intervention or mitigation actions by the government to address pandemic.
Communication efforts should aim at minimizing public confusion because no one knows the source of all these frustratingly obfuscated data. Everyone is trying to take a jab at making heads or tails of a never ending saga of spikes and waves in a country that has been in the longest lockdown in the world.
Understandably, not all countries will behave the same way in terms of number of confirmed cases and death rates. But to stay in the same trajectory for confirmed cases since the beginning of the ECQ tells us a different story altogether.
Trajectory of several countries in days since the confirmed cases first reached 30 per day. (The Philippines is in pink.)
Daily new confirmed COVID-19 deaths. (The Philippines is in pink.)
Daily new confirmed COVID-19 tests. (The Philippines is in pink.)
The figures above show the rolling seven-day average for confirmed cases, deaths, and testing done in various countries. (The Philippines is in pink.) Notice that while there are more tests being conducted in the Philippines over the past weeks, the trajectory shows that we have more or less maintained a plateau in daily confirmed cases. The conjecture that the ECQ was helpful in not increasing the numbers, while plausible, is also questionable. If the ECQ was indeed helpful, why did the number of confirmed cases not see a decline?
The answer—the backlog.
While there is a noticeable bending of the curve in deaths, there still remains a significant proportion of deaths, recoveries, and test results that have yet to be reported.
The health agency has also been latent in updating its own website and yet being able to come up with their “Beat COVID-19 Today” with apparently “real-time” information. One can’t help but ask: where did that information come from? When we decide to cherry pick the kind of news we want to dish out, one cannot help question the data.
The figure above shows the shift in reporting of the health agency a few weeks ago. Instead of reporting the confirmed cases daily, they attempt to report cases by date of onset of illness. (When did the patients get sick?)
They do this in order to “correct” for the delayed results. If you look closely at the figure, one cannot help wonder how 124 people in the BJMP in Cebu City got sick all on the same day (same as the BJMP in Mandaue, Barangay Mambaling, a spate of cases in the NCR and various barangays in Cebu).
Let’s give the benefit of the doubt to the data crunchers. If indeed the use of Date of Onset of Illness will provide us the “true” picture of COVID-19 cases, then logically since the clinical course will average 14 days then all patients before May 12 should have either recovered or died. Even given the benefit that half would have taken at least one month to fully recover, that would mean that all the patients from May 1 backward should have either recovered or died. Why is our recovery rate a sprinkling of 3,412 (out of 14,669 cases) or 23.2 percent? As of May 1, the DoH reported a total of 8,772. Why are the recoveries only 3,412 to date?
Given the fact that recovery rates are not better indicators over death rates in a pandemic, the point here is that accuracy and integrity in data is paramount in data analytics.
You can circumvent any information the way you want the outcomes to look. Good data will always be able to stand scrutiny anytime because no one needs a daily “explainer” to tell us why the numbers are the way they are.
For the day…
The announced new cases, recoveries, and deaths are tally of reported cases for the day.
Case fatality is lower at 6.04 percent (vs 6.21 percent global average, vs. 2.42 percent ASEAN average) and recovery rate is a tad up at 23.3 percent (vs. 42.1 percent world average, vs. 46.7 percent ASEAN average) for the day.
While we aim to target more than 10,000 tests a day, we’ve actually gone down in numbers for patients being tested. On May 25, the figure below shows that 307,434 total tests were conducted. This is only an additional 5,757 tests from May 24. Much lower than the seven-day average of 8,000 it achieved the previous week.
To say that we are controlling the pandemic will require the fulfillment of two parameters. The number of confirmed cases and the number of deaths must take the same trajectory. As long as there are active cases in the community, there will be ongoing transmission of the virus and the potential to cause disease. While there are less deaths noted, this may mean a more prepared healthcare system or a younger segment of the population being affected hence having better outcomes or a latency in the data on deaths. With continued expanded targeted testing, we should be able to identify those who are positive, perform aggressive contact tracing among those the patient has come in contact with, and isolate/quarantine them in order to curtail the chain of infection.
The number of posthumous results is 469. This means this is the number of people who have died before they were declared positive and is about 52.9 percent of all reported deaths.
New cases per day
Of the 350 new cases announced today, the DoH has not uploaded any information as to the region or city these “new cases” were reported from.
To say that it is challenging to cull accurate data from the agency is an understatement. We’re not after the good news if there is none to declare. The Filipinos are among the most resilient in times of crisis. But spare us the despondency of having to go through an emotional rollercoaster ride because seriously, many are asking the same question: based on the data provided, are we really ready for GCQ?
We all know that the pandemic is a worldwide problem. All we are asking is what’s the real score at home? The public deserves the right to know.
To get in touch with the Department of Health, the COVID hotline is (02)894-COVID loc 1555.
Disclaimers on the data:
(1) These three parameters (new confirmed cases, new recoveries, new deaths) are not real-time data. The data provided by the Department of Health is the date of public announcement. Even global data will vary in time of reporting, depending on the capacity of that country. To date, the latency period of the Department of Health on reporting recoveries averages almost 10 days (with more than 50 percent reported after eight days and more) and deaths averaging almost 9.5 days (with around 46 percent being reported after eight days or more).
(2) Depending on where testing is done, RT-PCR test results take an average of one to two days to process. Barring any delays, all tests done should ideally be released within 48 hours (the earlier the better). However, the test results released from government facilities range from three to 14 days, probably due to an overwhelming number of tests being conducted when compared to private hospitals.
TOTAL CONFIRMED CASES: 5,606,697
TOTAL DEATHS: 348,237 (case fatality rate: 6.21 percent)
TOTAL RECOVERED: 2,382,342 (case recovery rate: 42.1 percent)
Note that every reference has its own cut-off time for reporting. For the global data, WorldOMeters is used as its reference.
The total confirmed cases is close to the 5.5 million mark. The trend in the past week has begun to average 120,000 new confirmed cases daily with more testing being done worldwide. At the current growth rate, at least one million new confirmed cases may be registered every eight to nine days. The six million mark is projected to be breached on or before May 30, 2020.
Daily confirmed cases since December 31, 2019. The last time we looked back at the lowest numbers was on February 24, 2020. It has been an upward trajectory since and has plateaued at >80,000 confirmed cases per day since April 5.
The United States of America continue to lead globally in the number of total confirmed cases at 1,706,226 with a case fatality rate (CFR) of 5.85 percent with 99,805 total deaths recorded. The recovery rate for the US is up at 27.2 percent. Among the states, New York leads with 372,494 total confirmed cases and 29,310 total deaths with a CFR of 7.9 percent.
You may also like:
- What the COVID-19 budget tracker tells us about government spending so far
- This guy created a budget tracker that traces where government funds for COVID-19 goes
- While NCR has the most cases, it has one of the fewest quarantine beds
- With 38 new COVID cases overnight, Davao’s growth rate shoots past national average
- The COVID response budget for PH is actually bigger than P275B—it’s P380B, says budget tracker
Brazil remains in second with a CFR of 6.24 percent. Russia is in third with a CFR of one percent. India now breaks into 10th spot with 146,208 total confirmed cases, but a low CFR of 2.9 percent.
As countries ramp up testing in various degrees, more new confirmed cases are being reported. This means that with more testing and aggressive contact tracing, we are now able to identify patients who are asymptomatic or pre-symptomatic and are potentially infective to the vulnerable population. This explains why, with more testing, we now see a better picture of the extent of the pandemic. With more testing, the death rates go down.
The median average of case fatality rates worldwide has declined further to 6.21 percent. (It was 6.29 percent yesterday.) For the past two to three months, around 80 to 90 percent of patients are either asymptomatic or have mild disease and recover unremarkably.
Recoveries far outnumber the deaths with a ratio of approximately 6.85:1. (The ratio of recoveries continues to increase, and will reassuringly do so, over the deaths because of increased testing.)