How digital twins could transform healthcare

Dig­i­tal twins are well known in engi­neer­ing and man­u­fac­tur­ing. How­ev­er, sci­en­tists are now deploy­ing the tech­nol­o­gy in the med­ical realm, mod­el­ling the effec­tive­ness of exist­ing drugs against new dis­eases. 

Sci­en­tists and med­ical researchers often seek to treat patients by exam­in­ing the effi­ca­cy of exist­ing drugs. That’s what hap­pened when Covid-19 first arrived. Glob­al ana­lyt­ics com­pa­ny Else­vi­er and ExactCure, a start­up pro­vid­ing soft­ware solu­tions to reduce the impact of inac­cu­rate med­ica­tion, came togeth­er in April 2020 to build a dig­i­tal twin that could mod­el patient reac­tions to 20 approved drugs. 

But what is a dig­i­tal twin? In essence, they’re vir­tu­al rep­re­sen­ta­tions and sim­u­la­tions of a phys­i­cal asset or enti­ty. For exam­ple, phar­ma­ceu­ti­cal giant GSK has pilot­ed a repli­ca of its vac­cine pro­duc­tion process to improve oper­a­tions and future vac­cine devel­op­ment. 

The dig­i­tal twin from Else­vi­er and ExactCure was designed to sim­u­late the impact indi­vid­ual drugs would have on the bod­ies of patients with cer­tain phys­i­cal, genet­ic and med­ical char­ac­ter­is­tics. The data gleaned from the dig­i­tal twin could be used by physi­cians to per­son­alise future treat­ment plans and dos­es for patients infect­ed with SARS CoV‑2, the virus that caus­es Covid-19.  

Much like engi­neers use dig­i­tal twins to safe­ly test the impact of main­te­nance work, the tech­nol­o­gy could allow physi­cians to make clin­i­cal deci­sions in a safe envi­ron­ment. 

“One of the biggest ben­e­fits of dig­i­tal twins is the speed at which they can enable us to mod­el reac­tions, with­out involv­ing humans or ani­mals,” explains Dr Olivi­er Bar­ber­an, direc­tor of trans­la­tion­al med­i­cine solu­tions at Else­vi­er. This means that when there’s a pub­lic health cri­sis, “we can use data already avail­able, with no risk to life, to under­stand the dis­ease and how it inter­acts with already approved drugs to start fight­ing it as quick­ly as pos­si­ble.”

Patient power

Under­stand­ing the impact of a par­tic­u­lar drug on a patient before it’s admin­is­tered could ensure bet­ter care, out­comes and qual­i­ty of life, says Dr Frédéric Dayan, co-founder and CEO of ExactCure.

Health­care pro­fes­sion­als will begin to rely on sim­u­lat­ed and visu­al mod­els of treat­ment plans over the next decade or so, accord­ing to Dayan. “We will see dig­i­tal twins move from an R&D tool to one that empow­ers patients,” he adds.

Dig­i­tal twins could also be used to elim­i­nate – or at least reduce – the need for place­bos in clin­i­cal tri­als. Often when there’s no stan­dard treat­ment for a dis­ease, patients might be giv­en a place­bo made from a harm­less sub­stance like sug­ar or starch. While place­bos are con­sid­ered an impor­tant part of clin­i­cal research, they’re “eth­i­cal­ly ques­tion­able,” says Dr Gen Li, founder and CEO of Phe­si, a vir­tu­al clin­i­cal tri­als com­pa­ny. 

“The need to remove the use of place­bos is espe­cial­ly rel­e­vant when patients are in chron­ic and severe pain or where there’s enough his­tor­i­cal place­bo data from pre­vi­ous tri­als,” he adds. 

By using dig­i­tal twins, med­ical researchers could define the type of patients that should be includ­ed in a clin­i­cal tri­al and those that should be exclud­ed based on cer­tain char­ac­ter­is­tics. They could then pre­dict out­comes and side effects in real patients. The end result would be tri­als with few­er par­tic­i­pants.

“Work­ing towards erad­i­cat­ing the use of place­bos and reduc­ing patient bur­den should be an indus­try-wide goal,” Li argues. “Dig­i­tal twins will help us to reach this goal faster and we expect to see con­sid­er­able advances in the next five to 10 years.”

A question of trust

There’s no doubt­ing the life-chang­ing poten­tial of dig­i­tal twins. The Liv­ing Heart Project, led by Das­sault Sys­tèmes, was launched in 2014 to devel­op the world’s first mod­el of a human heart. Health­care spe­cial­ists, reg­u­la­to­ry bod­ies and med­ical device man­u­fac­tur­ers are work­ing togeth­er to boost the effec­tive­ness of car­dio­vas­cu­lar treat­ment. 

Part­ner­ships between tech­nol­o­gy com­pa­nies and the pub­lic health­care sec­tor will be cru­cial if dig­i­tal twins are to be adopt­ed more wide­ly. Some NHS trusts are already part­ner­ing with sim­u­la­tion spe­cial­ists to imple­ment dig­i­tal twins in bed man­age­ment. Man­u­al patient flows are often paper-based and inef­fi­cient; fore­cast­ing demand for beds enables health­care work­ers to opti­mise capac­i­ty and reduce wait times.

But bud­gets and resources are stretched for most hos­pi­tals. The more press­ing mat­ter for hos­pi­tals right now is tack­ling the back­log of peo­ple wait­ing for check ups, tests and scans.  Rishi Sunak, the chan­cel­lor of the exche­quer, pledged £5.9bn to the NHS in the Octo­ber bud­get to accel­er­ate the use of dig­i­tal tech­nol­o­gy in health­care set­tings and to reduce wait­ing lists. 

Then there’s the issue of pub­lic trust. Research from the Infor­ma­tion Commissioner’s Office, pub­lished in June, found that three in four peo­ple have high trust and con­fi­dence in the NHS. How­ev­er, 47% would be com­fort­able with the NHS shar­ing health data with pub­lic sec­tor com­pa­nies if it were to improve treat­ment, while only 42% would be hap­py for data to be shared with pri­vate sec­tor com­pa­nies. 

Every hos­pi­tal could one day have a dig­i­tal twin for every patient to treat any dis­ease, but this “seems like a fan­ta­sy and is unlike­ly to come to fruition in the next 20 to 30 years,” Dayan stress­es. For it to hap­pen, all hos­pi­tals would need to be imple­ment­ing dig­i­tal twins and shar­ing data to not only per­son­alise treat­ment and care plans but to advance the appli­ca­tions of dig­i­tal twin tech­nol­o­gy. This would require full pub­lic trust.  

“We don’t yet have the qual­i­ty of data to address each vari­able in the entire body to cre­ate an accu­rate full body dig­i­tal twin,” says Dayan. “But once we have enough accu­rate and reli­able data, it could become a real­i­ty.”