The miss­ing link be­tween pa­tient sur­veys and pos­i­tive out­comes

Modern Healthcare - - FRONT PAGE - Me­lanie Evans and Mau­reen Mckinney

The move­ment to im­prove hospi­tal qual­ity has long as­sumed that close ad­her­ence to best med­i­cal prac­tices leads to bet­ter out­comes. But a Mod­ern Health­care anal­y­sis of Medi­care’s first stabs at re­ward­ing hos­pi­tals that do well on process-of-care mea­sures and a pa­tient-sat­is­fac­tion sur­vey while pe­nal­iz­ing hos­pi­tals with high read­mis­sion rates shows only a mi­nor cor­re­la­tion be­tween the two.

The anal­y­sis, which uses Medi­care’s ini­tial re­sults for the value-based pur­chas­ing and read­mis­sions re­duc­tion pro­grams, also shows that hun­dreds of hos­pi­tals that did well in one pro­gram were pe­nal­ized in the other.

Dozens of hos­pi­tals that got the high­est marks for avoid­ing read­mis­sions were heav­ily pe­nal­ized for their per­for­mance on process and pa­tient sat­is­fac­tion, while dozens more that re­ceived the max­i­mum penalty for high 30-day read­mis­sion rates got high marks on key pro­cesses and at­ten­tion to pa­tients.

The faint cor­re­la­tion be­tween the two pro- grams un­der­scores the lim­its of the cur­rent qual­ity mea­sures, which have been cen­tral to the in­ten­si­fy­ing fed­eral ef­fort to re­duce health­care waste and er­rors and slow U.S. spend­ing on med­i­cal care, ex­perts say.

“The sci­ence isn’t quite there yet” to iden­tify which or­ga­ni­za­tional pro­cesses will de­liver de­sired out­comes such as lower read­mis­sion rates, said Dr. Rachel Werner, an as­so­ciate pro­fes­sor of medicine at the Univer­sity of Penn­syl­va­nia.

While re­searchers are work­ing to col­lect data on read­mis­sion-re­duc­tion strate­gies, she said, mea­sures have not yet been devel­oped for pro­grams that bet­ter co­or­di­nate med­i­ca­tions, doc­tors’ vis­its and med­i­cal care in set­tings out­side the hospi­tal. The CMS, mean­while, has fi­nal­ized plans to in­clude more out­comes-based mea­sures, such as risk-ad­justed mor­tal­ity, in fu­ture it­er­a­tions of its pay-for­per­for­mance pro­gram.

The CMS an­nounced its first round of re­wards and penal­ties for value-based pur­chas­ing, or VBP, in De­cem­ber. It scored hos­pi­tals based on their per­for­mance on clin­i­cal pro­cesses such as how quickly a heart at­tack pa­tient re­ceives an an­gio­plasty or whether a heart fail­ure pa­tient re­ceives dis­charge in­struc­tions. The fi­nal VBP score, which raised or re­duced Medi­care re­im­burse­ments by as much as 1%, also in­cluded a com­pos­ite mea­sure of pa­tient ex­pe­ri­ence.

The read­mis­sions re­duc­tion pro­gram, whose first round of penal­ties of up to 1% were an­nounced last fall, pe­nal­izes hos­pi­tals with higher-than-ex­pected num­bers of quickly re­turn­ing pa­tients. Pa­tients who re­turn to the hospi­tal within 30 days of a dis­charge are widely con­sid­ered a flag for lapses in care that may have caused their re­turn. Fed­eral penal­ties are set to steadily in­crease in coming years for ex­cess read­mis­sions, which are anath­ema to pa­tients, hos­pi­tals and pol­i­cy­mak­ers.

The Mod­ern Health­care anal­y­sis, which looked at the nearly 3,000 hos­pi­tals that were graded by both Medi­care qual­ity pro­grams, found only a weak like­li­hood that hos­pi­tals would do well in both.

The weak cor­re­la­tion can be il­lus­trated by the fact that hos­pi­tals that would be ex­pected to earn penal­ties in both pro­grams if the mea­sures were closely linked did not. More than 40% of hos­pi­tals that faced the max­i­mum 1% penalty for read­mis­sions earned bonuses for val­ue­based pur­chas­ing scores on pa­tient-sat­is­fac­tion and process-of-care mea­sures, a Mod­ern Health­care anal­y­sis of CMS data shows.

Mean­while, a sim­i­lar per­cent­age of hos­pi­tals that did not face penal­ties for ex­cess read­mis­sions none­the­less paid penal­ties for per­for­mance on value-based pur­chas­ing mea­sures. And among hos­pi­tals that earned some penal­ties on read­mis­sions, the re­sults were more evenly split for value-based pur­chas­ing bonuses and penal­ties.

Data used in the Mod­ern Health­care anal­y­sis does have some lim­its. The pop­u­la­tion varies some­what be­tween Medi­care’s two in­cen­tive pro­grams. Only the per­for­mance mea­sures are weighted for hospi­tal size. And the two in­cen­tive pro­grams draw data from dif­fer­ent time pe­ri­ods to cal­cu­late per­for­mance.

One likely rea­son for the weak cor­re­la­tion be­tween the two pro­grams is that the 12 clin­i­cal process mea­sures used by Medi­care’s VBP pro­gram do not cap­ture ev­ery­thing hos­pi­tals do to pro­mote high qual­ity and pre­vent re­peated hospi­tal vis­its, some ex­perts say.

“There is lim­ited ev­i­dence about the ef­fec­tive treat­ment and strate­gies for most acute­care con­di­tions,” said Dr. Mi­haela Ste­fan, an as­sis­tant pro­fes­sor of medicine at Tufts Univer­sity.

She re­cently led a study team that an­a­lyzed hospi­tal scores on more than two dozen process-of-care qual­ity mea­sures and read­mis­sion rates for pa­tients older than 65. The re­sults, pub­lished in the Jour­nal of Gen­eral In­ter­nal Medicine in Oc­to­ber, found hos­pi­tals with bet­ter marks for process-of-care mea­sures did not see lower rates of read­mis­sions than hos­pi­tals with worse marks. Still, Ste­fan said re­port­ing process mea­sures should con­tinue de­spite the weak cor­re­la­tion with read­mis­sions, since process-of-care mea­sures are among a lim­ited num­ber of qual­ity mea­sures with some ev­i­dence to show they do ben­e­fit pa­tients. Not ev­ery­one agrees. Dr. Ashish Jha, an as­so­ciate pro­fes­sor of health pol­icy and man­age­ment at the Har­vard School of Pub­lic Health, said early ef­forts to mea­sure qual­ity cor­rectly re­ported on pro­cesses of care. But the more valid mea­sure now is ac­tual out­comes, which is of most con­cern to pa­tients. “It’s time to move on,” Jha said.

Dis­con­nected care?

Oth­ers pointed to the pos­si­ble dis­con­nects be­tween what goes on within the hospi­tal and sub­se­quent care. The rate at which pa­tients re­turn to the hospi­tal within 30 days mea­sures a sys­tem of health­care that ex­tends be­yond the hospi­tal, said Arnold Ep­stein, a pro­fes­sor and chair­man of the Har­vard School of Pub­lic Health’s de­part­ment of health pol­icy and man­age­ment. Many fac­tors other than hospi­tal care can prompt a read­mis­sion.

For hos­pi­tals that have fared well un­der one pro­gram yet floun­dered un­der the other, the re­sults are con­found­ing.

MidMichi­gan Med­i­cal Cen­ter-Mid­land, a 250-bed hospi­tal, was deemed a top per­former by the CMS in curb­ing read­mis­sions and re­ceived no penalty un­der that pro­gram. Hospi­tal of­fi­cials credit the or­ga­ni­za­tion’s success to its par­tic­i­pa­tion in sev­eral vol­un­tary qual­ity im­prove­ment ini­tia­tives, in­clud­ing ones led by the state’s hospi­tal as­so­ci­a­tion and the In­sti­tute for Health­care Im­prove­ment.

But MidMichi­gan Med­i­cal Cen­ter-Mid­land was also among 1,423 hos­pi­tals that will lose up to 1% of their Medi­care pay­ments for 2013 un­der the CMS’ value-based pur­chas­ing pro­gram. The hospi­tal will have its pay­ments docked by 0.71%, ac­cord­ing to fed­eral data.

“You would ex­pect to see a cor­re­la­tion be­tween the two, but in our case, you just don’t,” said Kay Wag­ner, di­rec­tor of qual­ity for four-hospi­tal MidMichi­gan Health, the par­ent health sys­tem, also in Mid­land. “Our read­mis­sion rates are great but our process-of-care met­rics for heart fail­ure, for ex­am­ple, are not.”

Wag­ner ar­gued that the hospi­tal’s dif­fer­ent ex­pe­ri­ences in the two pro­grams il­lus­trate that qual­ity mea­sures—par­tic­u­larly those in use in fed­eral pro­grams—are still in their in­fancy. The dozen process-of-care mea­sures that ac­count for 70% of each hospi­tal’s score un­der value-based pur­chas­ing may not be the best pre­dic­tors of pa­tient out­comes, she said.

The re­main­ing 30% of a hospi­tal’s VBP score is based on a com­pos­ite mea­sure of the Hospi­tal Con­sumer As­sess­ment of Health­care Providers and Sys­tems, or HC­AHPS. The sur­vey, an eight-item in­stru­ment, col­lects pa­tient-re­ported data on ex­pe­ri­ence-re­lated vari­ables such as hospi­tal clean­li­ness, pain man­age­ment and com­mu­ni­ca­tion with physi-

cians and nurses.

MidMichi­gan Med­i­cal Cen­ter-Mid­land was in the midst of a large-scale ren­o­va­tion dur­ing the base­line pe­riod for 2013’s VBP pro­gram, which ran from July 1, 2009, un­til March 31, 2010. Those ren­o­va­tions—and the dust and noise that ac­com­pa­nied them—likely af­fected the hospi­tal’s HC­AHPS scores, Wag­ner said.

Still, Wag­ner backed the pro­grams and said the fed­eral government’s ad­di­tion of out­comes mea­sures, such as risk-ad­justed mor­tal­ity, in later ver­sions of value-based pur­chas­ing will likely make a big im­prove­ment. “They’re go­ing to learn by trial and er­ror,” she said.

Ukiah (Calif.) Val­ley Med­i­cal Cen­ter also ended up as a high scorer in the read­mis­sions pro­gram, but fell into the penalty range for the CMS’ pay-for-per­for­mance ini­tia­tive. The 62bed hospi­tal re­ceived no read­mis­sion penalty but will lose 0.65% of its Medi­care re­im­burse- ment in 2013 be­cause of its score on val­ue­based pur­chas­ing, ac­cord­ing to CMS data.

‘It’s a good be­gin­ning’

De­spite that in­con­sis­tency, hospi­tal of­fi­cials said the read­mis­sions and VBP pro­grams are good ini­tial gauges of hospi­tal qual­ity. “They might not be di­rectly cor­re­lated with each other right now, but it’s a good be­gin­ning and I know they’ll con­tinue to im­prove,” said Heather Van Housen, Ukiah Val­ley Med­i­cal Cen­ter’s vice pres­i­dent for pa­tient care. “I do think the cor­re­la­tion will be­come tighter later on.”

Michael Motte, CEO of 189-bed St. Alex­ius Hospi­tal, St. Louis, said he wasn’t sur­prised by the dif­fer­ence be­tween the hospi­tal’s per­for­mance in the value-based pur­chas­ing and read­mis­sions pro­grams.

Like MidMichi­gan Med­i­cal Cen­ter-Mid­land and Ukiah Val­ley, St. Alex­ius faces no read­mis­sions penalty, a success Motte cred­ited to close col­lab­o­ra­tion with lo­cal nurs­ing homes.

But St. Alex­ius Hospi­tal will see a 0.66% cut in its Medi­care re­im­burse­ment this year, a penalty that will cost the hospi­tal ap­prox­i­mately $81,000, Motte says.

“They’re two dif­fer­ent pro­grams fo­cus­ing on dif­fer­ent things,” he said of the hospi­tal’s con­trast­ing per­for­mance. “They’re go­ing to have dif­fer­ent re­sults.”

Value-based pur­chas­ing is based on older data, Motte said, and the hospi­tal has since en­acted a num­ber of changes that should boost its fu­ture score. They in­clude lead­er­ship changes, im­ple­men­ta­tion of health in­for­ma­tion tech­nol­ogy sys­tems and new ini­tia­tives aimed at in­creas­ing pa­tient sat­is­fac­tion.

“In my opin­ion, th­ese are good met­rics,” he said, “and I think we’re mov­ing in the right di­rec­tion.”

Source: CMS, Kaiser Health news

Source: CMS, Mod­ern Health­care anal­y­sis

Newspapers in English

Newspapers from USA

© PressReader. All rights reserved.