Putting costs on the map

Pop­u­la­tion’s health might trump geog­ra­phy: re­port

Modern Healthcare - - The Week In Healthcare - Jen­nifer Lubell

n as­sess­ing geo­graphic dif­fer­ences in Medi­care spend­ing, it may be wise to as­sess a pop­u­la­tion’s health rather than just the qual­ity of care it’s re­ceiv­ing, a study has con­cluded.

The anal­y­sis found that as­sess­ing ad­di­tional and bet­ter mea­sures of health re­duced the mag­ni­tude of un­ex­plained geo­graphic dif­fer­ences in Medi­care spend­ing, ac­cord­ing to re­sults pub­lished in the May 12 New Eng­land Jour­nal of Medicine and au­thored by re­searchers from Ge­orge Ma­son Uni­ver­sity and the Ur­ban In­sti­tute.

The trend in health­care re­form is to make providers more ac­count­able for the care they pro­vide by bundling pay­ments of hos­pi­tals and doc­tors, es­tab­lish­ing ac­count­able care or­ga­ni­za­tions, and pe­nal­iz­ing hos­pi­tals with high read­mis­sion rates (May 17, p. 6). How­ever, the NEJM study makes the point that “poli­cies that fo­cus on area-level spend­ing with­out ad­e­quate ad­just­ment for dif­fer­ences in ben­e­fi­cia­ries’ health sta­tus could in­ap­pro­pri­ately re­ward or pe­nal­ize cer­tain geo­graphic ar­eas,” said Jack Hadley, pro­fes­sor in the Col­lege of Health and Hu­man Ser­vices’ Health Ad­min­is­tra­tion and Pol­icy Depart­ment at Ge­orge Ma­son and se­nior author of the study, in an in­ter­view.

With­out know­ing all of the fac­tors that ac­count for geo­graphic dif­fer­ences in Medi­care spend­ing, pol­i­cy­mak­ers should re­sist look­ing for sim­ple so­lu­tions to limit ex­pen­di­tures in high-cost ar­eas, the study stated.

For hos­pi­tals in ar­eas clas­si­fied as “high cost” un­der Medi­care, “such in­sti­tu­tions should be en­cour­ag­ing their rep­re­sen­ta­tives to ar­gue in Congress that the cost dif­fer­ences across geo­graphic ar­eas aren’t nec­es­sar­ily due to in­ef­fi­ciency, and that they shouldn’t be get­ting pe­nal­ized by Medi­care,” Hadley said.

That’s not to say that in­ef­fi­cien­cies don’t ex­ist, Hadley said. Pa­tients may stay in hos­pi­tals when it isn’t nec­es­sary, or get ad­di­tional test­ing that isn’t needed. “It’s just that the geo­graphic dif­fer­ences in spend­ing may not be the best in­di­ca­tors of where those in­ef­fi­cien­cies are.”

Oth­ers in the in­dus­try who as­sessed the re­port’s find­ings agree that provider be­hav­ior,

Iwhile im­por­tant, doesn’t re­flect the en­tire story of cost vari­a­tion. “Geo­graphic vari­a­tion in ed­u­ca­tion lev­els, house­hold in­come, in­surance sta­tus prior to Medi­care en­roll­ment, smok­ing, ge­net­ics, etc., are all fac­tors in a com­plex equa­tion that re­sults in dif­fer­ences in dis­ease bur­den, ad­her­ence, pa­tient pref­er­ences, and ul­ti­mately, global uti­liza­tion,” said Wil­liam Golden, pro­fes­sor of medicine and pub­lic health at the Uni­ver­sity of Arkansas for Med­i­cal Sci­ences.

Pre­vi­ous stud­ies have of­ten iden­ti­fied vari­a­tion across geo­graphic ar­eas in how much Medi­care spends per ben­e­fi­ciary, Hadley stated. What they’ve found is “in the high­est-cost ar­eas Medi­care spends about 50% more per ben­e­fi­ciary than in the low­est-cost ar­eas, even af­ter ad­just­ing for ba­sic de­mo­graph­ics such as age, gen­der and race, and dif­fer­ences in med­i­cal-care prac­tices,” he said.

Re­gional dis­par­i­ties in Medi­care spend­ing was high­lighted last year in an ar­ti­cle in the New Yorker by sur­geon and author Atul Gawande, who found that higher spend­ing ar­eas weren’t nec­es­sar­ily de­liv­er­ing the best care, de­bunk­ing the claim that pa­tients get what they pay for. In par­tic­u­lar, he ex­plored the rea­sons why Medi­care spends twice the na­tional av­er­age per en­rollee in McAllen, Texas—one of the most ex­pen­sive health­care mar­kets in the coun­try—than in other com­pa­ra­ble mar­kets, even though the qual­ity of care there has been un­der scru­tiny (Nov. 9, 2009, p. 12).

For the NEJM study, how­ever, re­searchers took a spe­cific fo­cus on mea­sur­ing health, us­ing an ex­panded set of cri­te­ria that in­cluded both pre-ex­ist­ing con­di­tions and changes in health through­out the year. The study also con­trolled for pa­tients’ de­mo­graphic char­ac­ter­is­tics, fam­ily in­come, sup­ple­men­tal in­surance cov­er­age and area-level mea­sures of health­care sup­ply.

Re­searchers es­ti­mated the dif­fer­ences in Medi­care spend­ing be­tween high-and low­cost geo­graphic ar­eas, us­ing data on Medi­care spend­ing by 6,725 el­derly Medi­care pa­tients col­lected by the Medi­care Cur­rent Ben­e­fi­ciary Sur­vey from 2000 to 2002. Un­ad­justed, Medi­care spend­ing per ben­e­fi­ciary was 52% higher in the most ex­pen­sive re­gions than in the least ex­pen­sive re­gions. But af­ter ad­just­ing for de­mo­graphic and base­line health char­ac­ter­is­tics and changes in health sta­tus, re­searchers saw the dif­fer­ence in spend­ing be­tween the high­est and low­est spend­ing re­gions shrink to 33%.

“Our study shows that an in­di­vid­ual’s health ex­plains al­most onethird of the dif­fer­ence in Medi­care spend­ing per ben­e­fi­ciary be­tween the high­est-and low­est-cost ar­eas, while pre­vi­ous stud­ies have as­signed a smaller role to health mea­sures,” Stephen Zuck­er­man, a se­nior fel­low in the Ur­ban In­sti­tute’s Health Pol­icy Cen­ter and the lead author of the study, said in a writ­ten state­ment.

The prob­lem is that health sta­tus ex­plains only about 40% of these geo­graphic spend­ing vari­a­tions. “We’re still left with sub­stan­tial un­ex­plained vari­a­tions that likely deal with uti­liza­tion and pay­ment. The bot­tom line is we just don’t use health ser­vices in a con­sis­tent way,” said J.B. Sil­vers, a pro­fes­sor of healthsys­tems man­age­ment at Case Western Re­serve Uni­ver­sity in Cleve­land.

Sil­vers: “We just don’t use health ser­vices in a con­sis­tent way.”

Hadley: “Geo­graphic dif­fer­ences ... may not be best in­di­ca­tors.”

Newspapers in English

Newspapers from USA

© PressReader. All rights reserved.