Why dig­i­tal health­care is a balanc­ing act

José Mussi ex­am­ines the clin­i­cal ef­fec­tive­ness of new health­care tech­nolo­gies

Gulf Business - - CONTENTS - José Mussi

THERE ARE FEW sec­tors where the no­tion of hu­man in­ter­ac­tion seems so in­trin­sic as health­care. When we pic­ture it in our minds, the im­age that ap­pears isn’t a ma­chine: It’s a doc­tor, a nurse and a sur­geon. A calm, kind, re­as­sur­ing face.

Now imag­ine a health­care sec­tor of a dif­fer­ent kind, where most of our in­ter­ac­tion is not with a hu­man, but with a smart com­puter. In this sce­nario, our dis­cus­sions are with an in­tel­li­gent vir­tual agent, to which we ex­plain our prob­lems, symp­toms, and con­cerns, and from which we re­ceive our di­ag­no­sis and treat­ment plan. When we give blood, or un­der­take a med­i­cal pro­ce­dure or even surgery, we are in the hands of a ro­botic de­vice.

If it sounds like a vi­sion from the fu­ture, think again. This sce­nario is not just on the hori­zon; it is al­ready here, made pos­si­ble by the ex­po­nen­tial tech­nolo­gies in the sphere of health­care that have emerged over the last few years. At the cen­tre of this un­prece­dented change in one of our most piv­otal and beloved sec­tors are new so­lu­tions that in­cor­po­rate vary­ing lev­els of ar­ti­fi­cial in­tel­li­gence and ro­bot­ics that have left the lab and are now be­ing put into prac­tice across the world.

True, they can hardly be called com­mon­place yet. But make no mis­take, they will be.

How­ever, the in­evitabil­ity of change does not equate to us be­ing stripped of the abil­ity to in­flu­ence how this change takes place. It may be nat­u­ral for health­care to be em­brac­ing the dig­i­tal trans­for­ma­tion, given the im­pact it has al­ready had on other in­dus­tries. It may be an ob­vi­ous move to take full ad­van­tage of AI and ro­bot­ics to keep us healthy. But that doesn’t draw a line un­der the con­ver­sa­tion about pre­cisely how these new tech­nolo­gies will be im­ple­mented.

Of course, the philo­soph­i­cal de­bate about tech­nol­ogy re­plac­ing hu­mans and its so­cial im­pact is al­ready swirling and will con­tinue to do so. But that is not the only de­bate in town when we are talk­ing about a tech­no­log­i­cal trans­for­ma­tion in health­care. In many ways, it is not even the most im­por­tant one. In­stead, we need to fo­cus on whether the tech­nol­ogy is ready for use in the first place.

The ques­tions we asked our­selves at a re­cent con­fer­ence cen­tred on whether there is a way we can be sure these so­lu­tions are safe; whether they are more ef­fec­tive than, or even equally ef­fec­tive as, the tra­di­tional hu­man ap­proach to health­care; and how we can adapt ex­ist­ing ev­i­dence­based stud­ies to keep pace with in­no­va­tion.

Dig­i­tal health so­lu­tions are be­ing fo­cused on for one rea­son: they have proven ben­e­fits. There are many ex­am­ples that demon­strate the ca­pac­ity of AI-based di­ag­no­sis to be sev­eral times faster than hu­man in­ter­ven­tion. They make ad­vanced knowl­edge avail­able in stan­dard­ised form across the world, in­clud­ing in de­vel­op­ing coun­tries that would oth­er­wise not have ac­cess to this in­sight. Stud­ies have shown that the vast data­base of med­i­cal records can im­prove test­ing and di­ag­no­sis as com­pared to ex­ist­ing meth­ods.

But the ex­cit­ing and trans­for­ma­tional na­ture of tech­nol­ogy does not re­move the need for a balanced assessment of how it will be used and whether it is com­pletely safe. ‘New’ does not mean ‘per­fect’. And we still face the risk that many of the early re­sults and stud­ies men­tioned just now may be lim­ited to their spe­cific sce­nar­ios. And for all the pos­i­tive feed­back, not all of them have en­sured they are fully com­pli­ant with the pro­cesses and de­mands of full clin­i­cal tri­als.

Mak­ing mat­ters more com­pli­cated is the fact that AI is of­ten based on al­go­rithms de­signed to evolve over time – such as ma­chine learn­ing sce­nar­ios. We can have con­tin­u­ous tri­als of these math­e­mat­i­cal rules ev­ery time a new gen­er­a­tion of them emerges. But are we re­ally pre­pared to do that?

If tech­nol­ogy is to change the face of health­care in a way that makes the sec­tor more ef­fi­cient, ef­fec­tive, and stream­lined, en­abling it to help more peo­ple faster, sim­ply rush­ing it into oper­a­tion is not the an­swer.

AI should help users and doc­tors make bet­ter de­ci­sions, rather than mak­ing them it­self; the im­prove­ment of clin­i­cal ef­fec­tive­ness through so­lu­tions that have com­pleted full clin­i­cal tri­als should be just as much of a pri­or­ity as reg­u­la­tory and eth­i­cal con­sid­er­a­tions; and more work needs to be done to en­sure that so­lu­tions can be adopted across dif­fer­ent re­gions.

There must be recog­ni­tion of the fact that in­di­vid­ual pop­u­la­tions’ de­mo­graphic, so­cial, and ge­netic pro­files will im­pact how in­tel­li­gent so­lu­tions carry out their anal­y­sis, and for this to be taken into ac­count in the de­vel­op­ment and de­ploy­ment of these so­lu­tions. In health­care, there is no such thing as one-size-fits-all tech­nol­ogy.

Dig­i­tal health is al­ready here, and pretty soon it will be ev­ery­where. It is un­de­ni­able that health­care, and those it serves, can ben­e­fit im­mensely from it. But this is one sec­tor where, if a tech­nol­ogy is ‘ahead of its time’, it is rea­son for cau­tion rather than ex­cite­ment. The stakes are too high for it to be any dif­fer­ent.

José Mussi Se­nior di­rec­tor for dig­i­tal health and in­no­va­tion at PwC

Newspapers in English

Newspapers from UAE

© PressReader. All rights reserved.