What’s the right mea­sure of the im­pact of Covid-19?

Deaths cru­cial for un­der­stand­ing scale of cri­sis, gaps in re­sponse

Hindustan Times (Jalandhar) - - SPOTLIGHT - K Srinath Reddy and Surabhi Pandey The au­thors are with the Pub­lic Health Foun­da­tion of In­dia, or PHFI. Views are per­sonal.

NEW DELHI: On May 3, the day bridg­ing the sec­ond and third phases of the lock­down, where did the dif­fer­ent states of In­dia stand with re­gard to deaths re­sult­ing from Covid-19? How do we de­cide which states are do­ing bet­ter?

Usu­ally, me­dia re­ports men­tion the to­tal num­ber of deaths, but th­ese are not ad­justed for the size or the age struc­ture of the pop­u­la­tion.

That makes com­par­i­son be­tween states dif­fi­cult. A dif­fer­ent num­ber which stands out even more promi­nently in the re­ports is the Case Fa­tal­ity Ra­tio (CFR), which refers to the deaths as a per­cent­age of the cases di­ag­nosed.

Which is the num­ber that pro­vides a bet­ter com­par­i­son of the states?

CFR has a clearly iden­ti­fi­able nu­mer­a­tor: deaths at­trib­uted to Covid-19. The de­nom­i­na­tor can be very vari­able, based on the num­ber of per­sons who tested pos­i­tive for the virus and thus clas­si­fied as cases. That, in turn, de­pends on the num­ber of tests per­formed.

All who are tested for the virus will not be pos­i­tive, but more the num­ber of tests per­formed, higher the to­tal num­ber of per­sons de­tected, even if the frac­tion of pos­i­tive tests re­mains con­stant. Quite of­ten, when large num­bers are tested, more mild cases turn up among the pos­i­tives and they are less likely to die. So, a high test­ing rate, with more cases de­tected and a greater pro­por­tion of milder cases, will lower the es­ti­mate of CFR. A lower test­ing rate, with tests usu­ally re­served for the more se­vere cases, will yield a smaller de­nom­i­na­tor of cases and, there­fore, a higher CFR. South Korea, which tested lib­er­ally, in­clud­ing sev­eral asymp­to­matic per­sons, had a low CFR be­cause of the large num­ber and lower com­pos­ite risk pro­file of the cases in­cluded in the de­nom­i­na­tor.

On the other hand, to­tal deaths in any state are a clearly de­fined num­ber at any given time, as­sum­ing that deaths are be­ing cor­rectly clas­si­fied as Covid or non-Covid. How­ever, the ab­so­lute num­ber of deaths will also vary be­tween states ac­cord­ing to the size of their pop­u­la­tions. If we stan­dard­ise that mea­sure and es­ti­mate deaths per one mil­lion pop­u­la­tion, we can com­pare dif­fer­ent states more mean­ing­fully. This fig­ure yields a bet­ter pic­ture of a state’s over­all performanc­e against the virus, cap­tur­ing the ben­e­fit from the pop­u­la­tion level con­tain­ment strat­egy as well as the clin­i­cal case man­age­ment im­pact. It is this broad­band pack“Em­pow­ered age of pub­lic health and clin­i­cal in­ter­ven­tions that has a com­bined ef­fect on low­er­ing deaths in any state. On the other hand, CFR usu­ally gives us in­for­ma­tion only about the clin­i­cal treat­ment ben­e­fit. Only when we get a mea­sure of all cases in­fected in a pop­u­la­tion -- se­vere, mild and asymp­to­matic in­cluded -- do we get an In­fec­tion Fa­tal­ity Ra­tio (IFR), which in­vari­ably is lower than CFR.

The two sta­tis­tics (CFR and deaths per mil­lion) also give very dif­fer­ent pic­tures about the states (Charts 1 and 2). If we look at CFR, Delhi (1.6 ) looks bet­ter than Pun­jab (2.6) or Jhark­hand (2.6), which have lower test­ing rates. On the other hand, Pun­jab has far fewer deaths per mil­lion pop­u­la­tion (0.63) than Delhi (3.14). Jhark­hand is even bet­ter (0.08 deaths per mil­lion). Where would you rather be to­day? Megha­laya looks ter­ri­ble on CFR but is very well placed when the ac­tual death rate in the pop­u­la­tion is con­sid­ered. Odisha and Ker­ala do well in both, while Ma­ha­rash­tra and Gu­jarat stay high in both while min­i­mally chang­ing places .

Sta­tis­tics can be ap­peal­ing or ap­palling, de­pend­ing on how they are pre­sented. The pur­pose of each statis­tic dif­fers from another and the choice, of which to use when, is driven by the na­ture of as­sess­ment we wish to make. If performanc­e of hos­pi­tals is be­ing com­pared, CFR mat­ters. If the over­all state of the epi­demic is be­ing as­sessed, deaths per mil­lion is the statis­tic that tells us whether a state is in a safer zone than another. It is im­por­tant that all of us -- pol­i­cy­mak­ers, me­dia and the pub­lic -un­der­stand th­ese mea­sures and their dis­tinc­tions so that we can soon get the mea­sure of Covid-19.

If we look at case fa­tal­ity ra­tio, Delhi (1.6) looks bet­ter than Pun­jab (2.6) or Jhark­hand (2.6), which have lower test­ing rates.

On the other hand, Pun­jab has far fewer deaths per mil­lion pop­u­la­tion (0.63) than Delhi (3.14). Jhark­hand is even bet­ter (0.08 deaths per mil­lion).

K. SRINATH REDDY AND SURABHI PANDEY, Pub­lic Health Foun­da­tion of In­dia

Newspapers in English

Newspapers from India

© PressReader. All rights reserved.