On data-driven medicine

Modern Healthcare - - NEWS -

MH: In­ter­op­er­abil­ity pro­vides you with the abil­ity to co­or­di­nate care, and co­or­di­nat­ing care is all about the qual­ity, safety and, ul­ti­mately, the fi­nan­cial per­for­mance of your or­ga­ni­za­tions. Where are you on the jour­ney to build­ing the an­a­lyt­ics and us­ing data to sup­port care co­or­di­na­tion and de­liver bet­ter care?

Probst:

From a pop­u­la­tion health per­spec­tive, we’re on the first mile of our jour­ney. I think ev­ery­one is, whether it’s in the tech­nol­ogy un­der­ly­ing it or in un­der­stand­ing what we’re go­ing to do an­a­lyt­i­cally with that data to re­ally change health­care.

If you just ex­trap­o­late some of the best-prac­tice care pro­cesses that hap­pen in John’s or­ga­ni­za­tion, and the lives that are saved be­cause of that, the prob­lem is that the knowl­edge won’t trans­fer to Texas and Utah and other places. It’s pretty unique to what he can do in his or­ga­ni­za­tion. The abil­ity to share that knowl­edge is go­ing to save lots of lives.

Is that big data? It’s just kind of gen­eral an­a­lyt­ics and un­der­stand­ing how what we do to a pa­tient im­pacts that pa­tient, and be­ing able to share that knowl­edge. That’s what saves lives. Halamka:

I don’t know what big data is. I work on small data. It’s ac­tu­ally the new thing. (Laugh­ter)

How have I used that in clin­i­cal prac­tice? My wife was di­ag­nosed with Stage IIIA breast can­cer in De­cem­ber of 2011, Korean fe­male, Stage IIIA with HER-2 neg­a­tive, es­tro­gen-pos­i­tive pro­ges­terone mark­ers. The ques­tion I asked is, of the 10,000 pa­tients like her who have been seen in the past, how were they treated and what were their out­comes? We ac­tu­ally did that across 17 Har­vard hos­pi­tals, about 5 mil­lion pa­tients. We queried and found that Asian fe­males are very sen­si­tive to Taxol.

So we took the clin­i­cal trial data that said this dose of Taxol is good and di­vided it in half. She was treated, cured, and has no side ef­fects or any resid­ual is­sues. Why shouldn’t every pa­tient in Amer­ica in the fu­ture get this treat­ment? The tech­nol­ogy is there to­day.

Cham­bers: What’s a vi­able fi­nan­cial model for th­ese things to oc­cur? We can lament the death of the state health­care ex­changes, but as soon as the fed­eral fund­ing went away, they went away. There was no com­mer­cial vi­a­bil­ity to those mod­els.

The fi­nan­cially in­te­grated health­care provider is uniquely po­si­tioned to ad­dress a lot of th­ese is­sues … if we have per-mem­ber pre­mi­ums. The payer side of the equa­tion is the only way to fix this. If you look at huge dis­rup­tors in every place in the econ­omy, it is based on a per-month, per-mem­ber model. … That’s what health­care needs. We need to be on PMPM so we can get that pa­tient, make ad­just­ments, in­ter­act with their be­hav­ior and have a fi­nan­cial model where we can ac­tu­ally re­ward them as well.

MH: What about man­ag­ing pop­u­la­tions within your health sys­tems? Don’t you al­ready have the data from across the care con­tin­uum to do that?

Halamka: I’m from New Eng­land, and I am from your fu­ture. To­day, 70% of the in­come at all of our clin­i­cal sites are risk-based al­ter­na­tive pay­ment con­tracts with no fee-for-ser­vice. What do we have to do? It’s a to­tally dif­fer­ent IT. It’s not the trans­ac­tional bill. It’s fo­cused on qual­ity and out­comes.

We cen­tral­ize data into a sin­gle store across the com­mu­nity, and then a team of care man­agers en­roll, by dis­ease, in­di­vid­u­als into a guide­line or pro­to­col. We then man­age them as cus­tomers to de­ter­mine gaps in care and to en­sure com­pli­ance. We’ve gone be­yond

If you look at huge dis­rup­tors in every place in the econ­omy, it is based on a per-month, per-mem­ber model. … That’s what health­care needs.

Matthew Cham­bers

that in some dis­ease states like con­ges­tive heart fail­ure. We bought them all bath­room scales with Blue­tooth low-en­ergy in­ter­faces and get daily weights. If your weight goes up 7 pounds in a week­end, we dou­ble your Lasix (a diuretic) and take away your Dori­tos.

If you’re risk-based, you must treat pa­tients as a cus­tomer in their homes and in their com­mu­ni­ties. We are the No. 1 ac­count­able care or­ga­ni­za­tion in the coun­try and saved Medi­care $50 mil­lion last year.

Probst: We have been us­ing data to drive the de­ci­sions on how we should care for our pop­u­la­tions for a long time. If you have enough data, you can un­der­stand the best prac­tices to care for any given pop­u­la­tion. At In­ter­moun­tain, we’re do­ing a re­ally good job on chronic dis­ease be­cause we have a long his­tory of data that we can use. We’ve or­ga­nized our­selves around how to an­a­lyze that data, look at out­comes, and then im­prove that pop­u­la­tion.

We’re not as ag­gres­sive or un­der­stand yet how we can reach out to our pop­u­la­tion and take their Dori­tos away. But we’re on our way. We be­lieve in an­a­lyt­ics. Whether it’s big data, lit­tle data or what­ever kind of data, we be­lieve in hav­ing that data, be­ing able to an­a­lyze it, and then show­ing proof of out­come. We’ve been able to change be­hav­ior of physi­cians and peo­ple.

MH: How are the Dori­tos po­lice do­ing in Texas?

Cham­bers: I’m not from the fu­ture and if you lay your hand on an­other man’s Dori­tos in Texas, boy, I’m go­ing to tell you what. There’s a lot of con­cealed hand­gun car­ri­ers in the state of Texas. In all se­ri­ous­ness, per­sonal re­spon­si­bil­ity is the key. So you’re ob­vi­ously not ap­peal­ing to their brain. They know that they shouldn’t be do­ing it.

So you’re go­ing to ding them in the wal­let or ding them some­where. Some­body in our sys­tem said re­cently, “We’re ad­dicted to the crack of the fee-forser­vice model.” We know it’s go­ing away. The ques­tion is when and how quickly? It’s faster in New Eng­land, ap­par­ently.

We be­lieve in an­a­lyt­ics. Whether it’s big data, lit­tle data or what­ever kind of data, we be­lieve in hav­ing that data, be­ing able to an­a­lyze it, and then show­ing proof of out­come.

Marc Probst

Newspapers in English

Newspapers from USA

© PressReader. All rights reserved.