UBC Faculty Research and Publications

Impact of oncologist payment method on health care outcomes, costs, quality: a rapid review McPherson, Emily; Hedden, Lindsay; Regier, Dean A Sep 21, 2016

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-13643_2016_Article_341.pdf [ 809.43kB ]
JSON: 52383-1.0361990.json
JSON-LD: 52383-1.0361990-ld.json
RDF/XML (Pretty): 52383-1.0361990-rdf.xml
RDF/JSON: 52383-1.0361990-rdf.json
Turtle: 52383-1.0361990-turtle.txt
N-Triples: 52383-1.0361990-rdf-ntriples.txt
Original Record: 52383-1.0361990-source.json
Full Text

Full Text

RESEARCH Open AccessImpact of oncologist payment method onhealth care outcomes, costs, quality: a rapidreviewEmily McPherson1*, Lindsay Hedden2 and Dean A Regier1AbstractBackground: The incidence of cancer and the cost of its treatment continue to rise. The effect of these dual forces is amajor burden on the system of health care financing. One cost containment approach involves changing the wayphysicians are paid. Payers are testing reimbursement methods such as capitation and prospective payment while alsoevaluating how the changes impact health outcomes, resource utilization, and quality of care. The purpose of this study isto identify evidence related to physician payment methods’ impacts, with a focus on cancer control.Methods: We conducted a rapid review. This involved defining eligibility criteria, identifying a search strategy, performingstudy selection according to the eligibility criteria, and abstracting data from included studies. This process wasaccompanied by a gray literature search for special topics.Results: The incentives in fee-for-service payment systems generally lead to health care services being appliedinconsistently because providers practice independently with few systems in place for developing treatment protocolsand practice reviews. This inconsistency is pronounced in cancer care because much of the total per patient spendingoccurs in the last month of life. Some insurers are predicting that this variation can be reduced through the use ofprospective or bundled payments combined with decision support systems. Workload, recruitment, and retention are allaffected by changes to physician payment models; effects seem to be magnified in the specialist context as their severalextra years of training lower their overall supply.Conclusions: Experimentation with physician payment methods has tended to neglect cancer care providers.Policymakers designing cancer-focused physician reimbursement pilot programs should incorporate quality measurementsince very ill patients may receive too little treatment when payment models do not cover oncologists’ total costs, e.g.,fee-for-service systems whose prices do not account for the possible presence of other diseases.Keywords: Physician reimbursement, Physician payment, Oncology, Fee-for-service, Salary, Capitation, Activity-basedfunding, Prospective payment, Pay for performance, Payment by resultsBackgroundCancer is a leading cause of morbidity and mortality inCanada. In addition to human suffering, cancer annuallycosts the health system an estimated C$4 billion and isthe largest contributor to lost economic productivity [1].This trend will not abate, with incident cases in BritishColumbia (BC) expected to increase by 57 % between2012 and 2030 [2]. Providing high-quality care topatients that is also cost-effective is an ongoingchallenge for cancer control. Policymakers face the chal-lenge to control cost as increases in health care spendingput pressure on other government priorities such aseducation and defense. Since labor costs account for15 % of health care budgets in Canada [3], payers areexploring how they might contain costs by criticallyevaluating the way physicians are paid and how changesin payment method will affect health outcomes, resourceutilization, and quality of care.Physicians are tasked to deliver care that maximizes pa-tient benefit. Information asymmetry in medical treatment* Correspondence: mcpherson.emily@gmail.com1Canadian Centre for Applied Research in Cancer Control (ARCC), School ofPopulation and Public Health, University of British Columbia, 675 West 10thAvenue, Vancouver, British Columbia V5Z 1G1, CanadaFull list of author information is available at the end of the article© 2016 The Author(s). Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.McPherson et al. Systematic Reviews  (2016) 5:160 DOI 10.1186/s13643-016-0341-2requires principals (e.g., third-party payers; patients) torely on agents (e.g., physicians) to recommend and com-municate the consequences of alternative courses of ac-tion [4]. Agents must be incentivized to maximize benefitsto patients, rather than solely to the agent’s own benefit.In their most basic form, financial incentives from remu-neration are created through transferring money from theprincipals to the agent to provide care at a specified levelof quality. Economic theory suggests that incentives maybe used to reduce the marginal cost of physician behaviorchange, e.g., increasing adherence to evidence-basedguidelines [5]. If the size of the monetary incentive isgreater than the cost for the physician changing theirbehavior, the profit (or portion thereof) may be used as areward to the physician. The magnitude of change anddirection of behavioral response (i.e., incentive vs. disin-centive) will depend on a number of factors, including thecharacteristics of the incentive payment method and thefinancial and opportunity costs of participating in incen-tive schemes. These factors are important because poorlydesigned incentives may have unintended behavioral ef-fects and lead to lower levels of quality, e.g., if an overlylarge payment wrongly signals high risk [6].Major payment methodsThere are two primary attributes of physician remunerationthat influence the magnitude and direction of physician be-havior response: method of payment and level of payment[7]. Payment methods include capitation, fee-for-service,performance-based payment, prospective payment, and/orsalary. The timing of the payment can be prospective, i.e.,set in advance according to a fixed budget, or retrospectivewith or without a cap on total payments that are made.The second attribute, payment level, may be fixed in ad-vance or subject to negotiation after care is delivered. Al-ternatively, physicians may have complete or partialdiscretion as to the amount of money charged for services.The amount of payment for physician services may bereduced or withheld if behavior does not conform tobenefit-maximizing requirements. Finally, the amountmay vary depending on characteristics of the provider orpatients seen (e.g., more complex cases receive higher pay-ments). Table 1 provides an overview of each category ofpayment method, including the terms associated with thepayment approach, the definition of each category and thepotential benefits and harms of the payment approaches.In the United States (US), most recent reforms aimingto change the way health care is funded have focused onhospital payment, for example using global budgets,shared savings programs, penalties for readmissions, andhospital-acquired conditions, rather than changing theway providers are paid [8]. However, the 2015 MedicareAccess and CHIP Reauthorization Act offers Medicarepatients’ physicians a choice of payment models, e.g.,participating in the Merit-Based Incentive PaymentSystem, which starting in 2019 will adjust a provider’sfee-for-service reimbursement up or down based on pro-vider performance on quality measures that are currentlybeing developed [9]. In Canada, activity-based fundingfor hospitals have been implemented in at least threeprovinces, with most physicians receiving some form ofblended payment, followed by fee-for-service [10].Payment method impactsThe impacts of provider payment methods have beenextensively evaluated. For example, researchers havefound some correlation between the fee-for-servicepayment model and increased use of tests [11]. Theynote that higher service use does not necessarily im-prove outcomes and may even be harmful. Changesfrom fee-for-service to capitated payment models havealso been studied several times. This change does notappear to cause problematic decreases in primary careaccess [12], but does not decrease hospital use [13].The use of prospective payment may increase the rateof hospital readmissions and adverse events, but this in-crease is related to hospitals assigning more severe diag-noses to patients under prospective payment than theywould under other systems [14]. This means that thehospital is allocated more funding, although direct treat-ment costs have not increased.Cash bonuses have been demonstrated to improve someoutcomes—vaccination rates, for example—but researcherscaution that the improvement observed may be due to bet-ter reporting rather than true practice change [15]. The re-search indicates that combining payment model changeswith other interventions such as educational campaigns maybe needed to make meaningful practice changes [15].Provider payment reforms such as accountable careorganization (ACO) shared savings programs encourageproviders to form groups and assume responsibility for thecare of a population of patients in order to share in payersavings if quality and cost performance benchmarks areachieved. However, a 2001 study found that forming thesegroups had no significant effect on factors such as improvedcare coordination and innovation. The authors hypothesizedthat this may be because the groups did not identify as cohe-sive entities and used the structure mainly for legal purposesrather than to improve care provision [16].Broadly, existing evidence suggests that changes inphysician behaviors have the potential to impact the costand quality of care provided. Payment method can alsoinfluence recruitment and retention of physicians, whichin turn impacts patient access and quality of care [17].Object of the documentPrevious reviews [18–21] have evaluated the impact ofdifferent payment methods on cost and quality of care,McPherson et al. Systematic Reviews  (2016) 5:160 Page 2 of 15but none have focused specifically on the cancer carecontext. Across Canada, oncology remuneration takesseveral different forms. Sourced from the 2013 NationalPhysician Survey [22], Fig. 1 shows that salary is themost common remuneration method for medical oncol-ogists (37.7 % of respondents); other methods reportedare fee-for-service (17.8 %), “sessional/per diem” (3.9 %),none of these (1.2 %), and a blend of these methods(32.4 %). No oncologists reported payment by capitationor “incentives and premiums”.The objective of this study is to undertake a rapid re-view of the literature—a database search combined witha hand search of several systematic reviews and the grayliterature—to explore the impact of physician paymentmethods on patient outcomes, care quality, and overallexpenditure, with a specific focus on cancer control inCanada.MethodsWe followed guidelines for performing a rapid literaturereview [23], which included defining eligibility criteria,identifying a search strategy, performing study selectionaccording to the eligibility criteria, and abstracting data;this was followed up by an ancillary search for specialtopics. These searches were conducted in June 2015.Eligibility criteriaWe set out to include all studies published in English pub-lished in the past 10 years, regardless of whether they wereoriginal analyses or reviews of past work. We defined“impact” as the consequences of physician paymentmethods on health services use, expenditures, health out-comes, physician retention, and stakeholder opinion. Weexcluded abstracts, editorials, letters, and news.Search strategiesWe worked with a senior librarian at the BC CancerAgency to help identify subject headings and keywords.We also identified search terms and keywords from keybackground articles.Studies were identified through bibliographic searchesof the MEDLINE, Embase, and Evidence Based MedicineTable 1 Overview of physician payment approachesPayment model Definition Potential benefits and harmsCapitation; pre-payment Providers are paid a set amount for eachperson enrolled with them regardless ofwhether the person receives care.May reduce unnecessary health services utilization sincepayment is not tied to service provision. It is argued thefinancial incentives in capitation will lead primary carephysicians to reduce referrals to specialists [12]. However,some argue that providers may be incentivized to developoverly long lists and actually refer to specialist care toofrequently [33].Fee-for-service Providers are paid separately for allmedical services deliveredIn this method, providers are reimbursed for all medicalservices they provide, lowering the risk of taking on patientswho need many services. However, appointments may belimited to one service and complicated patients may requiremany appointments. This method may also increase the useof services which can give diminishing marginal returns oreven have detrimental effects [33] and incentivize theover-delivery of care because it rewards increasesin service volume, regardless of health benefit [11].Pay for performance; payment by results;performance-based payment; results-basedpurchasing; value-based purchasing;target paymentsProviders receive different payments formeeting or missing performancebenchmarks, e.g., related to quality,efficiency, care integration [8].Incentives based on achieving quality objectives areexpected to be associated with behaviors designedto achieve the quality targets, e.g., immunization rates,mammography screening, patient satisfaction scores [16].Risk adjustment algorithms should be employed sothat organizations are not penalized for treating sicker patients.Prospective payment; activity-basedfunding; bundled payment; lump-sumpayment; block funding; clinical pathwaysA fixed payment for each patient,based only on the patient’s diagnosisMay reduce clinical variation and end-of-life costs [40].However, without a focus on quality measurement, thepressures of these systems may place perverse incentiveson providers to deliver less care [51]. The developmentof “clinical pathways” (management plans that addressquality by providing the sequence and timing of actionscovered by the associated lump sum payment [55]) aimsto address this issue.Salary Individual providers get a fixed feeper year regardless of the numberof patients they treatSimilar to capitation, this method may have utilizationlowering effects. However, care quality may be compromised ifproviders respond to fixed payment by working shorter hoursand being less responsive to their patients’ needs anddemands [56].McPherson et al. Systematic Reviews  (2016) 5:160 Page 3 of 15Reviews publication databases using the following termsand variations upon these: Physicians Oncology Neoplasm Activity-based funding Prospective payment Bundled payment Lump-sum payment Pay for performance Payment by results Performance based payment Results-based purchasing Value-based purchasing Target payments Capitation Pre-payment Salary Fee-for-serviceUnder the guidance of the librarian, we undertook aniterative process to customize and refine the searchstrategy. Table 2 presents the full search strategies foreach database. Although the strategies contain state-ments that emphasize results specific to Canada, theseare combined with other statements that use “or”; assuch, the search has no regional limitation.Gray literature searchWe also undertook a gray literature search. This was lim-ited to oncology in Canada for time and scope reasons.We searched abstracts contained in the Canadian HealthHuman Resources Network online library [24], as well aswebsites for the following organizations: Canadian Association of Medical Oncologists [25] Canadian Association of Radiation Oncologists [26] Canadian Foundation for Healthcare Improvement [27] Canadian Institute for Health Information [28] Institute for Clinical Evaluative Sciences [29] National Physicians Survey [22] Statistics Canada [30]Study selection and data abstractionOne reviewer conducted screening. Initially, we reviewedarticle titles and abstracts for relevance; the full text of arti-cles that appeared to be potentially eligible were subse-quently reviewed for inclusion. EM abstracted the followingdata from each included article: authors, publication date,country, title, payment approach, health issue, outcomesmeasured, research methods, and study findings. Weworked closely with a senior librarian who helped designand calibrate the search strategies presented in full inTable 2. In-duplicate data extraction and post hoc data ex-traction review were not performed. We encourage futureFig. 1 Medical oncologist responses to 2013 National Physicians Survey [22] question 6a on remuneration method (percentage of total)McPherson et al. Systematic Reviews  (2016) 5:160 Page 4 of 15Table 2 Search strategies by databaseMEDLINE1 payment by results.mp. 1322 activity based funding.mp. 343 prospective payment.mp. 24714 results based purchasing.mp. 05 pay for performance.mp. 13886 value based purchasing.mp. 4067 performance based payment.mp. 328 Value-Based Purchasing/ 2219 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 437610 (salar* or cash or funding or remunerat*or reimburs* or capitation).m_titl.10,33111 exp reimbursement mechanisms/orexp fee-for-service plans/or expprospective payment system/19,94312 exp “Fees and Charges”/ 14,26513 economics, medical/or fees, medical/orexp Economics, Dental/411914 exp Income/ 30,58715 “costs and cost analysis”/ 17,76316 exp models, economic/ 941517 9 or 10 or 11 or 12 or 13 or 14 or 15 or 16 91,62018 Physician Incentive Plans/ 168619 exp Physicians/ec [Economics] 290120 economics, medical/or fees, medical/ 330821 Physician’s Practice Patterns/ 38,53222 “episode of care”/ 131423 Patient Care Bundles/ 5324 (physician* adj3 (remunerat* or reimburs*or payment*)).mp. [mp = title, abstract, originaltitle, name of substance word, subject headingword, keyword heading word, protocolsupplementary concept word, rare diseasesupplementary concept word, unique identifier]108225 exp reimbursement mechanisms/or expfee-for-service plans/or exp prospectivepayment system/19,94326 exp health personnel/ec 862227 18 or 19 or 20 or 21 or 22 or 23 or 24 or 25 or 26 70,14028 exp Neoplasms/ 1,465,61629 medical oncology/or radiation oncology/ 12,92130 exp Antineoplastic Agents/ 503,15431 Cancer Care Facilities/ 277232 Oncology Nursing/or Oncology Service, Hospital/ 618633 (cancer* or oncolog* or chemotherap*or radiotherap* or radiation therap*).m_titl.549,42934 28 or 29 or 30 or 31 or 32 or 33 1,800,02335 17 and 27 and 34 98336 exp Canada/ 77,345Table 2 Search strategies by database (Continued)37 (canad* or british columbia or albertaor ontario or quebec or manitoba orsaskatchewan or nova scotia or newbrunswick or newfoundland or princeedward island).mp. [mp = title, abstract,original title, name of substance word,subject heading word, keyword headingword, protocol supplementary conceptword, rare disease supplementary conceptword, unique identifier]118,54038 36 or 37 118,68039 27 and 34 and 38 28540 9 and (10 or 11 or 12 or 13 or 14 or 15 or 16)and 27 and 344741 9 and 27 and 34 5042 9 and 34 9943 9 and 27 and 38 3944 9 and 34 and 38 445 39 or 40 or 41 or 42 or 43 or 44 41946 limit 45 to yr = “2005 -Current” 28647 limit 46 to english language 275EmbaseEmbase <1974 to 2015 May 14># Search statement Results1 payment by results.mp. 2532 activity based funding.mp. 583 prospective payment.mp. 87244 results based purchasing.mp. 05 pay for performance.mp. 18596 value based purchasing.mp. 3207 performance based payment.mp. 438 Value-Based Purchasing/ 28109 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 13,74210 (salar* or cash or funding or remunerat*or reimburs* or capitation).m_titl.18,83511 exp reimbursement mechanisms/or expfee-for-service plans/or exp prospectivepayment system/55,98112 exp “Fees and Charges”/ 35,17113 economics, medical/or fees, medical/orexp Economics, Dental/651,97814 exp Income/ 71,06115 “costs and cost analysis”/ 53,72116 exp models, economic/ 115,18317 9 or 10 or 11 or 12 or 13 or 14 or 15 or 16 857,40918 Physician Incentive Plans/ 51,52219 exp Physicians/ec [Economics] 020 economics, medical/or fees, medical/ 45,65321 Physician’s Practice Patterns/ 183,69922 “episode of care”/ 209,403McPherson et al. Systematic Reviews  (2016) 5:160 Page 5 of 15reviews of this literature body that employ full systematicreview methods to include these activities.Study acquisition flowFigure 2 presents the acquisition flow of included studiesfrom the database search. Of 711 citations identified bythe database search, ten addressed the impact ofphysician payment methods on quality or access to care,equity, cost, or efficiency in the context of oncology. Al-though we are specifically interested in the effect of pay-ment methods in the Canadian context, the relativepaucity of studies encouraged us to include researchconducted outside Canada. Barring major contextual dif-ferences, payment method effects should be similaracross jurisdictions. Additional file 1 contains our popu-lated PRISMA checklist.ResultsThis section reviews the findings of the articles returnedby the database search and the ancillary search. A sub-section highlights the research methods used to evaluatethe consequences of physician payment methods.Table 2 Search strategies by database (Continued)23 Patient Care Bundles/ 18824 (physician* adj3 (remunerat* or reimburs*or payment*)).mp. [mp = title, abstract,subject headings, heading word, drugtrade name, original title, devicemanufacturer, drug manufacturer,device trade name, keyword]589625 exp reimbursement mechanisms/orexp fee-for-service plans/or expprospective payment system/55,98126 exp health personnel/ec 027 18 or 19 or 20 or 21 or 22 or 23or 24 or 25 or 26506,04428 exp Neoplasms/ 3,504,87929 medical oncology/or radiation oncology/ 108,56330 exp Antineoplastic Agents/ 1,565,42931 Cancer Care Facilities/ 19,45832 Oncology Nursing/or Oncology Service, Hospital/ 25,47633 (cancer* or oncolog* or chemotherap*or radiotherap* or radiation therap*).m_titl.1,058,60334 28 or 29 or 30 or 31 or 32 or 33 4,403,46335 17 and 27 and 34 12,76636 exp Canada/ 136,39837 (canad* or british columbia or albertaor ontario or quebec or manitoba orsaskatchewan or nova scotia or newbrunswick or newfoundland or princeedward island).mp. [mp = title, abstract,subject headings, heading word, drugtrade name, original title, devicemanufacturer, drug manufacturer,device trade name, keyword]228,07738 36 or 37 228,07739 27 and 34 and 38 164840 9 and (10 or 11 or 12 or 13 or 14or 15 or 16) and 27 and 3436741 9 and 27 and 34 38242 9 and 34 55443 9 and 27 and 38 11844 9 and 34 and 38 1445 39 or 40 or 41 or 42 or 43 or 44 230446 limit 45 to yr = “2005 -Current” 183347 limit 46 to english language 178848 “health policy economics and management”.ec. 504,36649 47 and 48 41650 (physician* and (fee or fees or pay*or remunerat* or compensat*or purchas* or reimburs*)).m_titl.237251 limit 50 to (english language andyr = “2005 -Current”)69452 49 or 51 110453 limit 52 to yr = “2013 -Current” 291Table 2 Search strategies by database (Continued)54 47 and physician*.mp. and (fee or feesor pay* or remunerat* or compensat*or purchas* or reimburs*).mp. [mp = title,abstract, subject headings, heading word,drug trade name, original title, devicemanufacturer, drug manufacturer,device trade name, keyword]15155 52 or 54 117756 limit 55 to exclude medline journals 8557 53 or 56 35258 remove duplicates from 57 342EBM ReviewsEBM Reviews - Cochrane Database ofSystematic Reviews <2005 to December 2014># Search statement Results1 (Cochrane Effective Practice and Organisationof Care Group).mp. [mp = title, short title,abstract, full text, keywords, caption text]912 1 and (canad* or british columbia).ti, kw. 03 1 and (cancer* or oncolog* or neoplasmsor tumor* or tumour* or chronic).ti,kw.34 1 and (physician* or specialit* or specialt*or dentist* or cost* or financ* or econom*or fees or reimburs* or pay* or salar* orremunerat* or fund* or cash or incentive*orbundle* or performance or capitation orpattern* or episode*).ti, kw.255 remunerat*.ti, kw. 16 3 or 4 or 5 28McPherson et al. Systematic Reviews  (2016) 5:160 Page 6 of 15Study characteristicsTable 3 presents the key characteristics of the databasesearch articles according to geographical location ofstudy, study design, sampling method, and sample size.The majority of included articles were from the US andthree studies were from Canada. Study designs includeda mix of quantitative and qualitative approaches, includ-ing observational studies (using administrative data-bases) [6, 12–14, 31, 32], literature reviews [11, 33–36],semi-structured interviews [37], and collection survey[15, 38–40]. Table 4 shows which payment methodseach article discusses.Articles identified from the database searchTable 5 summarizes the key attributes of the ten articlesidentified through the database search examining physicianremuneration in oncology. Four articles [11, 35, 36, 41] arebased on literature reviews or commentary. As such, thefindings may not be unique to the article.Habermann et al. [31] used data from the Medicarecancer registry (part of the National Cancer Institute’sSurveillance Epidemiology and End Results project) tocompare breast cancer screening rates in health main-tenance organizations (HMOs) with rates for providersreimbursed by fee-for-service. (Physicians practicing inHMOs are normally reimbursed through capitation[42]). Health maintenance organizations and managedcare generally tend to use more resources at the begin-ning of the care process, e.g., performing preventivecare, in order to keep people healthier and also savemoney in the long run [43]. Cancer stage at diagnosis, aproxy for screening rate, was estimated using a logisticregression model that adjusted for payment method aswell as demographic variables. The authors found thatwomen enrolled in health maintenance organizationswere more likely than those in fee-for-service to bediagnosed early (and therefore likely to have receivedscreening), both before and after a change from biennialto annual mammograms. However, this difference de-creased by half after the move to annual screening.Elit [36] analyzed the events that led to a change in pay-ment method for Ontario-based gynecologic oncologistsby conducting a non-systematic search of academic andgray literature and also speaking to key stakeholders in-cluding university physicians and members of the OntarioMedical Association. She found that most of theprovince’s gynecologic oncologists changed from fee-for-service remuneration to the salary-based program offeredin 2001 because they reported that prices in Ontario’s fee-for-service payment model did not account for the factthat non-oncology procedures are often more costly whenperformed on cancer patients. Study participants notedFig. 2 Study acquisition flow from database searchMcPherson et al. Systematic Reviews  (2016) 5:160 Page 7 of 15that the situation had encouraged specialists to focus onuncomplicated cases. Under-reimbursement had alsohampered the recruitment and retention of specializedstaff, but retention was improved when the alternativeprogram of salary payment was offered.In a related study, Elit and Cosby [37] conductedqualitative interviews which explored the impact ongynecologic oncologists of switching from fee-for-serviceto the alternative salary-based plan described above.They recruited 14 gynecologic oncologists from fivepractice sites in Ontario, four of which had opted for thechange from fee-for-service to salary. A semi-structuredinterview guide designed for the study was used in theinterviews; it consisted of professional and personalquestions. The interviewed physicians who experiencedthe remuneration change reported improvements intheir own quality of life and income predictability, whilealso noting that their preventive care work had in-creased. They noted that practice site vacancies wereable to be filled and staff were able to be retained. It hadbeen hoped that the new program would also reducephysician workload, but the interviewees stated this hadnot occurred. However, interviewees affirmed thatfollow-up for less complicated patients was now beingappropriately delegated.Newcomer et al. [44] describe a pilot project wherephysicians at five medical oncology groups in the USwere reimbursed with a single episode payment for allbreast, colon, and lung cancer patients at their initialvisit. All other physician services continued to be reim-bursed via the existing fee-for-service contract. Thestudy design compared the operational and control co-horts during the pre-pilot and pilot time periods. In theanalysis, 810 patients were used. Data included clinicaldata corresponding to characteristic of episode payments(cancer type, stage, genetic profile), the total medicalcost per episode of care (a linear regression function ofthe episode payment condition, age, and sex), andchemotherapy drug cost (average sale price). Controlswere obtained from UnitedHealthcare’s registry of morethan 65,000 breast, colon, and lung cancer patients. Thenet savings in total medical cost for the episode cohortcompared to fee-for-service was $33.4 million. Althoughthe program contained several incentives to lower drugcosts, chemotherapy drug spending unexpectedly rose; ittotaled $13.5 million more than predicted at $21 million.The authors state the study was not sufficiently poweredto analyze which expenses disproportionately impactedthe differences in total medical cost.Offering expert commentary on the Medicare pro-gram’s reimbursement for chemotherapy services, Bailesand Coleman [35] argue that Medicare’s fee-for-servicepayment system has tended to underestimate the totalcost of chemotherapy treatment. The authors state thatreimbursement for products and services used in the ad-ministration of chemotherapy drugs has often been sub-stantially less than their true cost. To cover theadministration resource shortfall, they note that oncolo-gists have relied on marginal profit from drug reim-bursement [45]. Decreases in drug payments brought inby the 2003 Medicare Modernization Act have resultedin losses for some oncologists. However, they note thatthe 2010 Affordable Care Act includes funding for pilotprograms to “align nationally recognized, evidence-basedguidelines of cancer care with payment incentives … inthe areas of treatment planning and follow-up care plan-ning” [46].Turning to pay for performance, Kuo et al. [32] con-ducted a retrospective analysis of breast cancer careexamining a program targeted at hospitals which re-wards better patient outcomes with a bundled payment,which encompasses treatment options based on recom-mended treatment plan for the breast cancer stage. Thispayment is higher than in the original payment scheme(case-based for surgery and fee-for-service for othertreatment components) when the treatment plan isfollowed, lower when it is not. The authors note that at-tending physicians in Taiwan are mainly employed byhospitals, so financial incentives applied at the hospitalTable 3 Characteristics of the identified articles from thedatabase searchCharacteristic Number of articlesGeographic locationCanada 2China 1Denmark 0Norway 0South Africa 0USA 7Study designLiterature review/commentary 4Qualitative survey/interviews 3Statistical analysis 3 (regression analysis)Sampling methodRandom 0All who agreed to participateand were eligible6Not applicable 4Sample size<100 3100 < n < 1000 1>1000 2Not applicable 4McPherson et al. Systematic Reviews  (2016) 5:160 Page 8 of 15level may still directly impact physician behavior. Datacame from the Taiwan Cancer Database. Women diag-nosed in 2003 or 2004 with stage I or II breast cancerwere included. The association of program participationand quality of care was estimated using linear regressionand controlling for age, stage, comorbidity, and type ofsurgery. Results showed that patients treated at hospitalsparticipating in the pay-for-performance program re-ceived higher-quality care, achieved better 5-year overallsurvival, and experienced less recurrence [30].Makari-Judson et al. [47] document the experience ofa group of 11 hematologic oncologists who were offeredperformance-based incentives in five categories (with as-sociated outcome in parentheses): patient-centered goals(measured by the patients’ medical record), quality mea-sures (Quality Oncology Practice Initiative metrics), clin-ical productivity (work relative value units), academic(not specified), and the group’s overall financial perform-ance (not specified). Incentives were arranged in threetiers; each corresponded to a category score andtriggered a bonus (percentage of salary). The authors re-port results for two of the five measures: patient-centered goals and quality measures. For the latter, “TierIII” was achieved resulting in a bonus of 24 %. For theformer, no bonus was achieved.Greenapple [40] conducted an online survey of 49American health insurers, representing more than 100million covered individuals, which asked them about themodels of care that they are implementing or wouldsupport in order to improve cancer care quality and alsocontrol cost. The survey results reveal that the payersmost favored systems of “clinical pathways”, a specializedform of care bundle where an evidence-based algorithmguides care practice for a defined group of patientsTable 4 Payment methods discussedArticle Capitation Fee-for-servicePerformance-based payProspective paymentSalary MixedBailes et al. 2014Bekelman et al. 2014Elit, L. 2006Elit et al. 2006GreenappleR. 2013Habermann et al.2007Kuo et al.2011Makari-Judson et al. 2013Newcomer et al. 2014Patel et al. 2013Total 2 (20%) 8 (80%) 4 (40%) 5 (50%) 1 (10%) 1 (10%)The shaded area indicates that the article in a given row discusses the payment method listed in the corresponding column headerMcPherson et al. Systematic Reviews  (2016) 5:160 Page 9 of 15Table 5 Database search articlesAuthors Title Payment approach Methods Health issue Outcomes measured FindingsBailes JS and ColemanTS. 2014 (USA)The long battle over paymentfor oncology services in theoffice setting [35]Fee-for-service Reviews Medicare policyhistory and reports expertopinionOutpatientChemotherapyPhysician fees forchemotherapy drugsPayments for drugadministration can bemuch less than its cost.Marginal revenue fromdrug payments is usedto make up the difference,and drug payment decreasescould result in provider losses.Bekelman JE, EpsteinAJ and Emanuel EJ.2014 (USA)Getting the next version ofpayment policy “right” on theroad toward accountable cancercare [11]Fee-for-service vs.prospective paymentReviews publishedliterature and agencydocumentationCancer care Changes in costsand outcomesProspective payment systemsshould include performancemeasurement to counterassociated perverseincentives. For complex caseslump sum payment could becombined with fee-for-service.Elit, L. 2006 (Canada) An analysis of alternativefunding for physicianspracticing gynecologiconcology in Ontario, Canadaprior to 2001 [36]Fee-for-service Literature search,discussion withstakeholders, meetingminutes from groupsconsidering alternatefunding systemsGynecologiccancerEvents preceding reformof the funding agreementwith gynecologiconcologistsFee-for-service does notaccount for the increasedcomplexity of services oncancer patients, causinglosses and makingrecruitment and retentiondifficult.Elit L, Cosby J andGynecologicOncology Group inOntario. 2006(Canada)Does shifting a physicianpayment system shift physicianpriorities? A multi-siteevaluation of an alternativepayment plan (APP) for gynecologiconcologists in Ontario [37]Fee-for-service vs. anegotiated arrangementwhere contracts are madewith physician groups whoare paid a fixed amountregardless of productivityInterviews with 14 Ontariogynecologic oncologists;interviewswere analyzed usinggrounded theory.GynecologiccancerChanges in physicianbehavior in responseto the new paymentsystemThe new plan improvedquality of life and incomepredictability, increasedpreventive health care work.Vacancies were filled andstaff were retained. Staffdelegated follow-up with lesscomplicated patients. Theplan did not reduce workload.Greenapple R. 2013(USA)Rapid expansion of new oncologycare delivery payment models:results from a payer survey [40]Comparing “clinical pathways”(bundled payments with qualitymanagement), capitation, sharedsavings and pay-for-performanceA validated survey ofpayers representing morethan 100 millionindividuals that askedpayers about models ofcare that could improvequality and reduce costs.Cancer care Payer perceptions ofwhich payment modelsare most effectivePayers believe that clinicalpathways can reduce clinicalvariation in care, improvequality and reduce costs,mainly by reducingend-of-life costsHabermann EB, VirnigBA, Riley GF, andBaxter NN. 2007 (USA)The Impact of a Change inMedicare Reimbursement Policyand HEDIS Measures on Stage atDiagnosis Among Medicare HMOand Fee-For-Service Female BreastCancer Patients [31]Fee-for-service vs. healthmaintenance organization(capitation)Compares the effect ofchange from biennial toannual mammograms bypayment method.Breast cancer Surveillance Epidemiologyand End Results, Medicareclaims databaseWomen enrolled in the healthmaintenance organizationwere more likely than thosein fee-for-service to bediagnosed early both beforeand after the, but after thechange, the disparity shrankfrom 4.7 to 2.3 %.McPhersonetal.SystematicReviews (2016) 5:160 Page10of15Table 5 Database search articles (Continued)Kuo RN, Chung KPand Lai MS. 2011(China)Effect of the pay-for-performanceprogram for breast cancer care inTaiwan [40]Fee-for-service/activity-basedfunding vs. pay-for-performance(encouraging evidence-basedtherapy and reward better pa-tient outcomes)A retrospective analysis ofpatients who receivedcurative surgery.Multivariate regressionanalyzed the associationbetween programenrollment and quality ofcare.Breast cancer Population-based cancerregistration and claimsdataEnrollees received higher-quality care, had better5-year overall survival andless recurrenceMakari-Judson G,Wrenn T, Mertens WC,Josephson G andStewart JA. 2014(USA)Using Quality Oncology PracticeInitiative Metrics for PhysicianIncentive CompensationPay for performance Based on theirperformance in fiveachievement categories,physicians were offered abonus percentage ofsalary corresponding tothe target level achieved.HematologyoncologyWork relative value units,Quality Oncology PracticeInitiative metrics, patientemotional well-beingfrom medical records,academic goals and theoverall financial successof the groupResults are reported for twomeasures: quality andemotional well-being. For theformer, “Tier III” was achievedresulting in a bonus of 24 %salary. For the latter no bonuswas achieved.Newcomer LN, GouldB, Page RD, DonelanSA and Perkins M.2014 (USA)Changing Physician Incentivesfor Affordable, Quality CancerCare: Results of an EpisodePayment ModelFee-for-service vs. episodepayments (bundled payments)Physicians at five medicaloncology groups werereimbursed with a singleepisode payment forservices to cancer patientsas part of a pilot program.The episode cohort wascompared with a controlfee-for-service cohort.Breast, colonand lungcancerClinical datacorresponding tocharacteristics of episodepayments (cancer type,stage, genetic profile),claims data, averagechemotherapy drugsale priceThe total medical cost for theepisode cohort was $33.4million less than what waspredicted using fee-for-service.Patel KK, Morin AJ,Nadel JL andMcClellan MB. 2013(USA)Meaningful Physician PaymentReform in OncologyClinical pathways (bundledpayments), pay for performance,fee-for-serviceReviews pilot initiatives inthe US that combinephysician paymentreforms with deliveryreforms.Cancer care Research on oncologypractice and the impactof physician paymentmethods, proposals fromoncology societiesThe authors propose apayment model thatcombines fee-for servicepayment with casemanagement payment and acare coordination fee,increasing total providerpayment but potentiallydecreasing the total care cost.McPhersonetal.SystematicReviews (2016) 5:160 Page11of15during a set period of time [45]. They believe that thesecould reduce both the cost of end-of-life care and clin-ical variation in care, while also improving care quality.The payers prioritize controlling costs through themethod of reducing wasteful and inappropriate care andbelieve that clinical pathways are most likely to achievesuch reductions.In an expert commentary, Patel et al. [41] review pilot ini-tiatives in the US that combine oncologist payment reformswith delivery reforms, including performance incentives,bundled payment and clinical pathways, and mixedmethods, i.e., a fee-for-service chemotherapy payment forthe cost of buying the drug, fixed payments for drug ad-ministration and care management. They propose a phys-ician payment model for cancer care that combines fee-for-service payment with case management payment (to lowerthe incentive to increase the volume and intensity of patientservices) and a care coordination fee. This would involveincreasing total payment to physicians but could decreasethe total cost of cancer care (by decreasing waste and ineffi-ciency, as well as payments for all other cancer care).Bekelman et al. [11] use academic literature andpublic-sector publications to make evidence-based rec-ommendations on reforms to cancer care payment pol-icy. They argue that any prospective payment systemsshould focus on performance measurement, since theorypredicts that lump sum payment systems will place per-verse incentives on providers, e.g., providing too few ser-vices to very ill patients [6]. The authors recommend astrategy of cross-subsidizing with fee-for-service com-plex cases treated under prospective payment in order tomitigate the risk to providers of having their total costsexceed the lump payment [11].Research methods used to evaluate consequences ofphysician payment methodsRegression analysis was used by two articles in the data-base search. One used a pretest-posttest study design[31], where one group is assessed at different timepoints. The other used a retrospective cohort analysis[32] where two groups are compared at the same time.Qualitative research methods were employed by twoarticles [37, 40]. Their study designs involved using lit-erature reviews to inform a semi-structured interviewguide or surveying payers and conducting in-person in-terviews with providers. To analyze the qualitative re-sults, Greenapple [40] calculated the percentagebreakdown of participant responses. Two investigatorsindependently analyzed coded data in Elit and Cosby’sstudy [37]. Then together they discussed themes and de-veloped theory; the model that emerged from the discus-sion was validated with two final interviews.Regarding the creditability of study results in general:the studies tended to be exploratory and not designed toprovide generalizeable results, but more than two thirdsof the articles performed at least some uncertainty ana-lysis of their results.Data sources used to evaluate consequences of physicianpayment methodsSince the provider compensation method was often usedat the outset as an indicator dividing the sample into co-horts or “treatment groups”, studies seldom includedpayment method as an exogenous variable. Instead studyauthors compared payment method cohorts using dataon utilization and expenditures [32] and on patient out-comes and stakeholder opinion (collected through in-person interviews and online surveys) [37, 40].Results from the gray literature searchMultiple sources provide information on the ways that on-cologists are paid in Canada. However, they do not investi-gate the outcomes associated with the different payments[10, 22, 29]. For example, the Canadian Institute forHealth Information’s (CIHI) National Physician Database2012–2013 [10] includes payment information for medicalspecialists, e.g., Fee-for-Service Clinical Payments toPhysicians by Province/Territory; oncology is one of thespecialties that comprise the medical specialist category.DiscussionThis review presents the literature’s key findings relatedto the impact of different physician payment methodsand uncovers articles that examine those impacts in thecontext of cancer care. This is important since the studyfindings show that payment method impacts on cancercare can run contrary to what would be expected inother disease areas. For example, although Ellis [6]showed that providers reimbursed via fee-for-service areunlikely to discriminate for patient illness severity, Elit[36] saw that oncologists reimbursed through fee-for-service methods were substituting toward patients withless complicated conditions since the fee levels were notsufficient for treating cancer patients.The Medicare program has been a leader in experiment-ing with models of physician payment, but it has not fo-cused on the area of cancer care [48]. Indeed this reviewappears to be one of the first treating the impact of phys-ician payment methods on cancer care. Its findingsemphasize the importance of further study of the impactsof changing payment methods for the physicians whofocus on cancer care. The gray literature search shows thatthe CIHI National Physician Database includes oncologistremuneration information by province in Canada andcould be used as data source for future projects.Cancer care tends to include high costs concentratedat the end of life, when relatively low-cost palliative caremay be a more effective option, both financially and withMcPherson et al. Systematic Reviews  (2016) 5:160 Page 12 of 15regard to the patient’s quality of life. It is estimated thatthe Medicare program spends one third of the cost oftreating cancer in the final year of a patient’s life and78 % of that spending occurs in the final month [49].Broomberg [33] argues that this is an expected result ofincentives in fee-for-service payment systems, whichreinforce doctors’ tendency to apply health care re-sources inconsistently as they practice independentlywith few systems in place for developing treatment pro-tocols and practice reviews. It follows that the payersqueried in Greenapple’s [40] survey believe that re-placing individual services with effective “bundles” of on-cology care could bring down end-of-life costs whilealso improving quality and reducing regional variation.Workload, recruitment, and retention are all affected bychanges to physician payment models. Effects seem to bemagnified in the specialist context as their several extrayears of training lowers their supply to the system. Pay-ment models that lead to poor retention of providers, forexample, if remuneration does not cover physician costsas in the case of chemotherapy drug administration ser-vices [35], may lead to heavy workloads which in turncomplicate recruitment. Also it has been noted that theproportion of patients assigned “high-severity” status foraccounting purposes can be significantly reduced whenthe workload of the discharging physician is increased,resulting in a substantial revenue loss for the hospital [50].However, few of the studies that focused on the specialistcontext examined workforce factors such as these as out-comes [37]. Clearly, future studies aiming to evaluate theimpact of changes to payment methods for oncologists inCanada should include measurement and analysis ofchanges in workload and workforce factors. Levels of hos-pital utilization and expenditures/claims are obvious start-ing points in areas where oncologists are primarilyemployed by hospitals. Yet in addition to these general in-dicators, payers may also be interested in changes in thelikelihood of adverse events since these can have direct,predictable impacts on utilization and expenditure; hos-pital readmission rates could be used to proxy for adverseevents. In terms of office-based oncologists, changes inthe number of tests ordered after patient consultationscould be examined. In both settings, conformity withguidelines for evidence-based medicine is a way to evalu-ate the impact of pay-for-performance payment systems.It is also crucial to evaluate how much of the variationobserved after a new payment method’s implementationshould be attributed to factors other than that method.A few studies took up this challenge. Kristiansen et al.[39] report that the variables used in their analyses ex-plain only 10 % of the observed variation in laboratoryutilization; as a result, they note that the main determin-ant of test ordering behavior is probably the medicalcondition. This emphasizes the necessity of conductingstudies in the cancer context: to discover factors thatdrive cost, but can be changed while controlling for un-changing disease complexity. It should also be notedthat two studies [14, 38] used random sampling whileothers targeted a specific population and then includedall or most members of the population who agreed toparticipate.Many payers are experimenting with different paymentstrategies. However, some study results may not be gen-eralizeable to other contexts, either because of the re-search methods or also the structure of the healthsystem studied. An example of structural difference isthe Taiwanese health system where doctors (includingthose at the primary care level) are almost all employedby hospitals [32]. Reforms instituted at the hospital levelmay broadly affect physician behavior, but this will notbe the case in systems where doctors practice independ-ently. Another example is in single payer context wherepayment models may need to include specific reimburse-ment for teaching and research services [36]. In multi-payer systems, having several funding streams may makethese activities more likely to be funded by one of thepayers. As experimentation continues, we should re-member that most studies report short-term effects ofpayment system changes, but the longer term-associatedchanges in technology use and practice structure may bemuch larger in magnitude [51]. For example Finkelstein[52] has shown that the implementation of Medicare in1965, which caused a large increase in fee-for-service re-imbursement, led to much larger effects on cost, tech-nology use, and practice delivery over time than wassuggested by initial, static analyses.This review experiences some methodological limita-tions common to rapid literature reviews. The lack of asystematic review process implies some study selectionbias. Although we evaluate the credibility of study find-ings, we did not include a systematic quality assessmentprocess; it has been argued that forgoing such a process isa source of bias in rapid reviews [53]. However, the bodyof literature was so small that eliminating work based on aquality rubric would likely have left us too little to review.Another possible limitation is that we did not perform in-duplicate data extraction and post hoc data extraction re-view. Further, in the current policy environment, where asnoted above the MACRA legislation is overhaulingphysician payment methods in the huge system of USMedicare, rapid reviews have an advantage over systematicreviews because they can be produced quickly to facilitateevidence-based policymaking.The review includes only articles published in English.While limiting results in this way could omit Quebec-specific results, prior research [54] shows that English-language-restricted literature searches tend to have similarresults to those without language restrictions (when theMcPherson et al. Systematic Reviews  (2016) 5:160 Page 13 of 15review’s content is mainly based within published litera-ture). A follow-up gray literature search could be expandedto include French content.ConclusionsThis review presents currently published literature re-lated to the impact physician payment method has oncancer care. It shows that, although general impacts ofphysician payment methods have been well-studied, re-search has seldom been extended to the specialized cir-cumstances of cancer care.Nevertheless, several findings have implications fordecision-makers concerned with the impact of physicianpayment systems on cancer care. Patients with high-severity illnesses may receive too little treatment in bun-dled payment systems that rely on patient diagnosis so itmay be prudent to invest in quality measurement pro-grams when implementing these systems. Fee-for-servicepayment models can also lead to too little treatment whenprices do not vary to account for patient status, e.g., whena treatment is not complex in principle, but it is made soby the overall poor health of the patient. However, we seethat even high levels of treatment are not necessarily acorollary for quality of care, so even when fee-for-serviceincentivizes the provision of too much treatment, it maystill need to incorporate quality measurement.Cancer care is resource-intensive: technologies are ex-pensive and treatments are time-intensive. The time isright to evaluate outcomes that occur before and afterreforms to oncologist payment methods, for example,BC’s recent move from fee-for-service to salary-basedpayment. Others could leverage the results of this ex-periment to avoid costly duplication. Recruitment andretention rates should also be examined to further quan-tify impacts of new programs; effects on research pro-duction and teaching programs are also of interest.Additional fileAdditional file 1: PRISMA checklist. (PDF 218 kb)AbbreviationsBC: British Columbia; CIHI: Canadian Institute for Health Information;HMO: Health maintenance organizations; US: United StatesAcknowledgementsThis review is sponsored by the Institute for Health System Transformation &Sustainability (IHSTS). IHSTS was not involved in the conduct of the review.The project was led by researchers from the School of Population and PublicHealth, University of British Columbia, the Faculty of Health Sciences, SimonFraser University, and the Canadian Centre for Applied Research in CancerControl (ARCC), BC Cancer Agency. ARCC is funded by the Canadian CancerSociety (Grant # 2015-703549). We thank Diana Hall of the BC Cancer Agencywho provided library services.Authors’ contributionsEM designed the search strategy, reviewed returned articles, appliedinclusion restrictions, and drafted the manuscript. DR and LH offeredcomments throughout the research and drafting process. All authors readand approved the final manuscript.Competing interestsThe authors declare that they have no competing interests.Author details1Canadian Centre for Applied Research in Cancer Control (ARCC), School ofPopulation and Public Health, University of British Columbia, 675 West 10thAvenue, Vancouver, British Columbia V5Z 1G1, Canada. 2Centre for ClinicalEpidemiology and Evaluation, University of British Columbia, 828 West 10thAvenue, Vancouver, British Columbia V5Z 1M9, Canada.Received: 9 November 2015 Accepted: 14 September 2016References1. Economic Burden of Illness in Canada, 2005-2008. Ottawa, Ontario: PublicHealth Agency of Canada; 2014.2. Woods RR, Coppes MJ, MBA AJC. Cancer incidence in British Columbiaexpected to grow by 57% from 2012 to 2030. 2012.3. Canadian Institute for Health Information. National Health ExpenditureTrends, 1975 to 2012. 2012.4. Arrow K. The economics of agency. In: Pratt JZ, editor. Principals and agents—thestructure of business. Boston: Harvard Business School Press; 1985. p. 37–51.5. Gneezy U, Meier S, Rey-Biel P. When and why incentives (don’t) work tomodify behavior. J Econ Perspect. 2011;25:191–209.6. Ellis RP. Creaming, skimping and dumping: provider competition on theintensive and extensive margins. J Health Econ. 1998;17:537–55.7. Brown C. Wage levels and method of pay: National Bureau of EconomicResearch. 1990.8. Schneider EC, Hussey PS, Schnyer C. Payment reform: analysis of modelsand performance measurement implications. Santa Monica, CA: RANDcorporation; 2011.9. Access M. CHIP Reauthorization Act of 2015. Pub L. 2015;16(114-10):129.10. National Physician Database, 2012-2013. Canadian Institute for HealthInformation. 2014.11. Bekelman JE, Epstein AJ, Emanuel EJ. Getting the next version of paymentpolicy “right” on the road toward accountable cancer care. Int J RadiatOncol Biol Phys. 2014;89(5):954–7. doi:10.1016/j.ijrobp.2014.04.022.12. Davidson SM, Manheim LM, Werner SM, Hohlen MM, Yudkowsky BK,Fleming GV. Prepayment with office-based physicians in publicly fundedprograms: results from the Children’s Medicaid Program. Pediatrics. 1992;89(4):761–7.13. Hutchison B, Birch S, Hurley J, Lomas J, Stratford-Devai F. Do physician-payment mechanisms affect hospital utilization? A study of Health ServiceOrganizations in Ontario. CMAJ. 1996;154(5):653–61.14. Cutler DM. The incidence of adverse medical outcomes under prospectivepayment. Econometrica. 1995;63(1):29–50.15. Fairbrother G, Hanson KL, Friedman S, Butts GC. The impact of physicianbonuses, enhanced fees, and feedback on childhood immunizationcoverage rates. Am J Public Health. 1999;89(2):171–5.16. Shortell SM, Zazzali JL, Burns LR, Alexander JA, Gillies RR, Budetti PP, et al.Implementing evidence-based medicine: the role of market pressures,compensation incentives, and culture in physician organizations. Med Care.2001;39(7):162–78.17. Physician recruitment improves care in northern Alberta. Grand Prairie:Alberta Health Services 2011 October 12. Report No.: 30.18. Gosden T, Forland F, Kristiansen IS, Sutton M, Leese B, Giuffrida A, et al.Capitation, salary, fee-for-service and mixed systems of payment: effects onthe behaviour of primary care physicians. Cochrane Database Syst Rev.2000;3, CD002215. doi:10.1002/14651858.CD002215.19. Gosden T, Pedersen L, Torgerson D. How should we pay doctors? Asystematic review of salary payments and their effect on doctor behaviour.QJM. 1999;92(1):47–55.20. Palmer KS, Agoritsas T, Martin D, Scott T, Mulla SM, Miller AP, et al. Activity-based funding of hospitals and its impact on mortality, readmission,discharge destination, severity of illness, and volume of care: a systematicreview and meta-analysis. PLoS ONE. 2014;9(10), e109975. doi:10.1371/journal.pone.0109975.McPherson et al. Systematic Reviews  (2016) 5:160 Page 14 of 1521. Van Herck P, De Smedt D, Annemans L, Remmen R, Rosenthal MB, SermeusW. Systematic review: effects, design choices, and context of pay-for-performance in health care. BMC Health Serv Res. 2010;10:247. doi:10.1186/1472-6963-10-247.22. National Physician Survey. http://nationalphysiciansurvey.ca/. Accessed 2June 2015.23. Ganann R, Ciliska D, Thomas H. Expediting systematic reviews: methods andimplications of rapid reviews. Implement Sci. 2010;5(1):56.24. Canadian Health Human Resources Network Library. http://tools.hhr-rhs.ca/index.php?option=com_mtree&Itemid=109&lang=en. Accessed 2 June 2015.25. Canadian Association of Medical Oncologists. http://cos.ca/camo/. Accessed2 June 2015.26. Canadian Association of Radiation Oncologists. http://www.caro-acro.ca/.Accessed 2 June 2015.27. Canadian Foundation for Healthcare Improvement. http://www.cfhi-fcass.ca/Home.aspx. Accessed 2 June 2015.28. Canadian Institute for Health Information. http://www.cihi.ca. Accessed 2June 2015.29. Institute for Clinical Evaluative Sciences. http://www.ices.on.ca/. Accessed 5June 2015.30. Statistics Canada. http://www.statcan.gc.ca/start-debut-eng.html. Accessed 2June 2015.31. Habermann EB, Virnig BA, Riley GF, Baxter NN. The impact of a change inMedicare reimbursement policy and HEDIS measures on stage at diagnosisamong Medicare HMO and fee-for-service female breast cancer patients.Med Care. 2007;45(8):761–6.32. Kuo RN, Chung KP, Lai MS. Effect of the Pay-for-Performance Program forBreast Cancer Care in Taiwan. J Oncol Pract. 2011;7(3S):e8s–e15s. doi:10.1200/JOP.2011.000314.33. Broomberg J, Price MR. The impact of the fee-for-service reimbursementsystem on the utilisation of health services. Part I. A review of thedeterminants of doctors’ practice patterns. SAMJ. 1990;78:130–2. doi:Review.34. Coulam RF, Gaumer GL. Medicare’s prospective payment system: a criticalappraisal. Health Care FinancRev. 1992;1991:45–77.35. Bailes JS, Coleman TS. The long battle over payment for oncology services inthe office setting. J Oncol Pract. 2014;10(1):1–4. doi:10.1200/JOP.2013.001272.36. Elit L. An analysis of alternative funding for physicians practicinggynecologic oncology in Ontario, Canada prior to 2001. Eur J GynaecolOncol. 2006;27(1):61–4.37. Elit L, Cosby J. Does shifting a physician payment system shift physicianpriorities? A multi-site evaluation of an alternative payment plan (APP) forgynecologic oncologists in Ontario. Eur J Gynaecol Oncol. 2006;27(4):375–8.38. Krasnik A, Groenewegen PP, Pedersen PA, Von Scholten P, Mooney G,Gottschau A, et al. Changing remuneration systems: effects on activity ingeneral practice. BMJ. 1990;300:1698–701.39. Kristiansen IS, Hjortdahl P. The general practitioner and laboratory utilization:why does it vary? Fam Pract. 1992;9(1):22–7.40. Greenapple R. Rapid expansion of new oncology care delivery payment models:results from a payer survey. Am Health Drug Benefits. 2013;6(5):249–56.41. Patel KK, Morin AJ, Nadel JL, McClellan MB. Meaningful physician paymentreform in oncology. J Oncol Pract. 2013;9(6S):49s–53s.42. Zuvekas SH, Cohen JW. Paying physicians by capitation: is the past nowprologue? Health Aff. 2010;29(9):1661–6.43. Sethi MK, Frist WH. an introduction to health policy: a primer for physiciansand medical students. Springer Science & Business Media. 2013.44. Newcomer LN, Gould B, Page RD, Donelan SA, Perkins M. Changingphysician incentives for affordable, quality cancer care: results of an episodepayment model. J Oncol Pract. 2014;10(5):322–6.45. ASCO in Action Brief: Payment Reform Models Explained. American Societyof Clinical Oncology 2013 March 15. Report No.: 4.46. Patient Protection and Affordable Care Act. 2010.47. Makari-Judson G, Wrenn T, Mertens WC, Josephson G, Stewart JA. Usingquality oncology practice initiative metrics for physician incentivecompensation. J Oncol Pract. 2014;10(1):58–62.48. Falit BP, Chernew MD, Mantz CA. Design and implementation of bundledpayment systems for cancer care and radiation therapy. Int J RadiationOncol Biol Phys. 2014;89(5):950–3.49. Hogan C, Lunney J, Gabel J, Lynn J. Medicare beneficiaries’ costs of care inthe last year of life. Health Aff. 2001;20(4):188–95.50. Powell A, Savin S, Savva N. Physician workload and hospital reimbursement:overworked physicians generate less revenue per patient. Manuf ServiceOper Manag. 2012;14(4):512–28.51. McClellan M. Reforming payments to healthcare providers: the key toslowing healthcare cost growth while improving quality? J Econ Perspect.2011;25:69–92.52. Finkelstein A. The aggregate effects of health insurance: evidence from theintroduction of Medicare: National Bureau of Economic Research. 2005.53. Juni P, Altman DG, Egger M. Systematic reviews in health care: assessing thequality of controlled clinical trials. BMJ. 2001;323:42–6.54. Egger M, Juni P, Bartlett C, Holenstein F, Sterne J. How important arecomprehensive literature searches and the assessment of trial quality insystematic reviews? Empirical study. Health Technol Assess. 2003;7:1.55. Panella M, Marchisio S, Di Stanislao F. Reducing clinical variations with clinicalpathways: do pathways work? Int J Qual Health Care. 2003;15(6):509–21.56. Rosen B. Professional reimbursement and professional behavior: emergingissues and research challenges. Soc Sci Med. 1989;29(3):455–62.•  We accept pre-submission inquiries •  Our selector tool helps you to find the most relevant journal•  We provide round the clock customer support •  Convenient online submission•  Thorough peer review•  Inclusion in PubMed and all major indexing services •  Maximum visibility for your researchSubmit your manuscript atwww.biomedcentral.com/submitSubmit your next manuscript to BioMed Central and we will help you at every step:McPherson et al. Systematic Reviews  (2016) 5:160 Page 15 of 15


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items