Health­care’s ‘money­ball’

Pre­dic­tive mod­el­ing be­ing tested in data-driven ef­fort to strike out hos­pi­tal read­mis­sions

Modern Healthcare - - SPECIAL REPORT - Melanie Evans

Hos­pi­tals wait­ing on Congress to squeeze spend­ing may sym­pa­thize with the Oak­land Ath­let­ics’ re­ver­sal of for­tune more than a decade ago, when the ball club’s own­ers re­fused to con­tinue prof­li­gate spend­ing. Soon there­after, as read­ers of the best­seller

Money­ball know, Oak­land A’s Gen­eral Man­ager Billy Beane re­lied on fancy math and eco­nomic the­ory to put to­gether a com­pet­i­tive base­ball team on a shoe­string ma­jor league bud­get.

Beane’s con­vic­tion—that sta­tis­ti­cal anal­y­sis could trump base­ball tra­di­tion that blinded the sport to valu­able play­ers and left Oc­to­ber to teams able to af­ford er­rors—be­came the sub­ject of Michael Lewis’ 2003 book and a newly re­leased movie.

“We take 50 guys and we cel­e­brate if two of them make it,” Beane fumes about the ma­jor league draft in Lewis’ book. “In what other busi­ness is 2-for-50 a suc­cess?”

Now hos­pi­tal ex­ec­u­tives, flooded with data from in­for­ma­tion tech­nol­ogy in­vest­ments and un­der pres­sure to curb waste and spend­ing, face the same ques­tion that Beane and oth­ers be­fore him, in­clud­ing Bill James, an in­flu­en­tial author and base­ball statis­ti­cian, sought to an­swer: What mea­sures mat­ter the most?

In health­care, the ques­tion is crit­i­cal when it comes to who ends up in the hos­pi­tal.

Hos­pi­tals, which house squadrons of nurses, pricey tech­nol­ogy, phar­ma­cies and lab­o­ra­to­ries, are costly places to care for pa­tients, with hos­pi­tal care ac­count­ing for health­care’s sin­gle largest ex­pense. Pol­i­cy­mak­ers see sig­nif­i­cant po­ten­tial to curb spend­ing by keep­ing more pa­tients out of the hos­pi­tal. In 2005, Medi­care paid hos­pi­tals $7,200, on av­er­age, and $12 bil­lion in to­tal for re­peat hos­pi­tal vis­its that could have been avoided, ac­cord­ing to one es­ti­mate by the Medi­care Pay­ment Ad­vi­sory Com­mis­sion.

Hos­pi­tals also can be un­safe. Pa­tients risk in­fec­tions or other avoid­able com­pli­ca­tions dur­ing a hos­pi­tal stay that, para­dox­i­cally, can land them back in the hos­pi­tal shortly af­ter leav­ing.

And hos­pi­tals now have added in­cen­tives to bet­ter iden­tify such pa­tients. Medi­care, the sin­gle largest cus­tomer for many hos­pi­tals, will pe­nal­ize those with too many re­peat pa­tients start­ing in 2013. The penalty starts at up to 1% of hos­pi­tal Medi­care rev­enue, then in­creases over the next sev­eral years to as much as 3%.

In­deed, within hos­pi­tals and health plans, some have started to use the same fancy math as the Oak­land A’s and other pre­dic­tive mod­els to iden­tify pa­tients at risk for un­nec­es­sary hos­pi­tal stays.

“Noth­ing good hap­pens when you go to the hos­pi­tal,” says John Billings, di­rec­tor of the Cen­ter for Health and Pub­lic Ser­vice Re­search at New York Univer­sity’s Robert F. Wag­ner Grad­u­ate School of Pub­lic Ser­vice. Billings and col­leagues de­vel­oped an al­go­rithm be­ing tested by Med­i­caid in New York. “Go­ing to the hos­pi­tal means some­thing went wrong out­side of the hos­pi­tal,” Billings says.

Hide­bound health­care?

Sim­i­lar­i­ties be­tween base­ball’s ad­her­ence to some­times in­ef­fec­tive and costly con­ven­tion and the in­ef­fi­ciency of U.S. health­care have not es­caped no­tice.

“Amer­ica’s health­care sys­tem be­haves like a hide­bound, tra­di­tion-based ball club that chases af­ter ag­ing slug­gers and plays by the old rules: We pay too much and get too lit­tle in re­turn,” wrote Beane and health pol­i­cy­mak­ers Newt Gin­grich and Sen. John Kerry (D-Mass.) in a 2008 New York Times editorial.

Newspapers in English

Newspapers from USA

© PressReader. All rights reserved.