CMAAO Coronavirus Facts and Myth Buster 123: Corona Prediction Models

943: Round Table Expert Zoom Meeting on “Corona prediction models”

 Formulas in relation to COVID-19 pandemic for better understanding of data and risk stratification

6th June, 2020, 11am-12pm

Participants: Dr KK Aggarwal, Dr DR Rai, Dr Suneela Garg, Dr Girdhar Gyani, Dr Narottam Puri, Prof Mahesh Verma, Dr Ashok Gupta, Dr JA Jayalal, Mrs Upasana Arora, Dr Jayakrishnan Alapet, Dr K Kalra, Mr Anil Kumar, Director HCFI, Dr Sanchita Sharma

Formulas in relation to COVID-19 pandemic for better understanding of data

  • Wuhan conducted 6.5 million tests (nasal swab, RT PCR) for coronavirus in 10 days, when 6 new cases were detected; 200 cases were found, mostly people who showed no symptoms. The ratio of undocumented cases for each documented case in Wuhan is 1:33, i.e., there were 33 asymptomatic cases for every one new infection. This ratio in New York was found to be 1:10, i.e., there were 10 asymptomatic patients, for every positive patient.
  • We need to know realistic figures as many will be presymptomatic or asymptomatic. Comparison of data should be between epicenter and epicenter, non-epicenter and non-epicenter.
  • Statistics help in preparedness.
  • Number of expected deathsis 15% of the number of serious patients 14 days back.
  • Death rateis the number of deaths today vs. number of cases today.
  • Corrected death rateis the number of deaths today vis a vis number of cases 14 days back.
  • Estimated number of deaths:For every one death, there will be one unreported death. It is calculated as number of reported deaths x 2.
  • Deaths are not labeled as due to COVID-19 because the primary cause of death is reported as, for example, pulmonary embolism; COVID-19 is reported as a secondary cause of death.
  • Number of tests per million population: More the number of tests, more will be the number of cases. Government guidelines advise against screening of asymptomatic cases.
  • Doubling rate:India 14 days, France 49 days; if there are 2.5 lakh cases in India today, this number will double in about 15 days to 5 lakhs and so on.
  • Seir model: It talks about how many are susceptible, how many are exposed, infected or recovered.
  • Non pharmacological interventions should be taken into consideration; at least 50% of quarantine will be helpful.
  • We need to stratify the impact of COVID-19 on hospitalization, ICU admissions, and fatality rate.
  • Of those who died, 67% had comorbidities, while 37% had no comorbidity. It is this group which should not be ignored.
  • There is similarity between dengue and COVID-19.

o   In dengue, when fever resolves, then capillary leakage may occur and patient may become critical. Administration of hydration at this point of time will reduce or prevent mortality.

o   In COVID-19, after fever, silent hypoxia occurs – this gives 5 hour window. At the time when hypoxia develops, if home oxygen is administered, first dose of LMWH is given and then search for hospital is started – this will buy 5-6 extra hours and may prevent mortality. A ventilator is not the answer. Early oxygen is the answer; early heparin is the answer.

  • It is important to calculate the number of people who will require oxygen. This data will help to reduce mortality.

o   Number of oxygen requirement is 10% of new admissions today; this will happen after 7 days. For example, if there are 1000 cases in Delhi today, 100 patients will need oxygen therapy after 7 days. But actually, there may be 30,000 asymptomatic cases (if the ratio from Wuhan study is applied) and 3000 of these will need oxygen after 7 days.

o   Number of ventilator requirement is number of cases today x 3, which will manifest on the 9th day.

  • According to ICMR guidelines for home discharge, mild, very mild and presymptomatic cases need not be tested at the time of discharge. But, presymptomatic cases cannot be diagnosed without testing.

Risk stratification – “Am I at risk?”

  • Male or female: Females are protected till menopause because of female sex hormones, then they are at similar risk as males.
  • Childrenup to 2 years are protected due to thymus gland. The hyperactive thymus produces cytokines, which are protective. In children, mortality is seen only in those with congenital heart disease, nephrotic syndrome.
  • Am I in a high-risk group?The chances of healthcare workers are highest in ICU and ER personnel, intensivists, ENT doctors, ophthalmologists and dentists.
  • What are my chances of acquiring infection in OPD? Follow the Singapore protocol.

o   If contact time is less than 10 minutes and the patient and the doctor have both used a mask (simple surgical mask and N95, respectively), the risk is very low.

o   If the contact time is less than 30 minutes, the risk is mild.

o   If the contact time is more than 30 minutes, the risk is high and you need to undergo test.

  • Can you tolerate hypoxia?If you have uncontrolled diabetes, heart failure, COPD, asthma, you will not be able to tolerate hypoxia even for 2 hours. You are at high risk.
  • Are you at risk of hypercoagulation?If you have inflammatory disease such as Crohn’s, rheumatoid arthritis, SLE, ulcerative colitis, Sjogren, you are vulnerable to hypercoagulable state despite anticoagulant. Such cases, even if asymptomatic, will suddenly develop embolism on Day 8/9, which may lead to death.
  • Surgical time:Keep surgical time less than 30 minutes.

If you take all these precautions, you are at controlled risk and can continue to work.

Being male and having any cardiac decompensated state increases your risk.  A cardiac patient cannot tolerate anoxia, tachycardia and irregular heart rate. All these three occur in COVID-19. Opt for teleconsultations.

  • All doctors should carry a portable pulse oximeter to detect silent hypoxia, which occurs on Day 7 even if you are asymptomatic.
  • Keep an oxygen concentrator in your office and at home. Monitor oxygen.

 

Dr KK Aggarwal

President CMAAO, HCFI and Past National President IMA

Be the first to comment

Leave a Reply

Your email address will not be published.


*