Scholarly article on topic 'Dispelling the myths around cancer care delivery: It's not all about costs'

Dispelling the myths around cancer care delivery: It's not all about costs Academic research paper on "Economics and business"

CC BY-NC-ND
0
0
Share paper
Academic journal
Journal of Cancer Policy
OECD Field of science
Keywords
{Cancer / Cost / Outcome / Adoption / Diffusion / Innovation / "Cancer drugs" / "Dynamic process" / "Clinical relevance" / Guidelines}

Abstract of research paper on Economics and business, author of scientific article — Carin A. Uyl-de Groot, Elisabeth G.E. de Vries, Jaap Verweij, Richard Sullivan

Abstract The costs of cancer care grow exponentially. It has been argued that there is a linear relation between costs and outcome: the more a country spends on cancer care, the better the outcome. We try to dispel this myth, by showing that the relation is not linear at all and by describing other factors in the cancer care delivery process that have an impact on outcome. We show that there is a correlation between health care expenditure and life expectancy at birth, but that there is no correlation between number of deaths per 100,000 and cost per person spent on cancer in general, neither in lung, breast, colorectal and prostate cancer. Furthermore, a decrease in survival can be related to accessibility, affordability or equity issues, but also to factors such as life style. In the real world the process of cancer delivery is complex and dynamic, with many (potential) innovations. When efficacy is proven and an innovation is considered clinically relevant, the innovation has to be incorporated in evidence based clinical guidelines. However, implementation in such a guideline is still no guarantee for optimal adoption and diffusion of an innovation. Cancer care delivery also goes beyond matters related to health-systems and cancer costs, new technologies, reimbursement agencies, hospitals, and health-care professionals by increasingly involving shared decision making. An optimal process of cancer care delivery consists of the use of new and existing diagnostic tests and treatment strategies of high quality and is effective, safe, patient centred, efficient and timely. Such health system is highly recommended and all stakeholders in society will benefit.

Academic research paper on topic "Dispelling the myths around cancer care delivery: It's not all about costs"

CANCER POLICY

g 1 <" , ^ Contents lists available at ScienceDirect

Jp|l | ® Journal of Cancer Policy

ELSEVIER journal homepagewww.elsevier.com/locate/jcpo

Review

Dispelling the myths around cancer care delivery: It's not all about costs^-^

Carin A. Uyl-de Groota*, Elisabeth G.E. de Vriesb, Jaap Verweijc, Richard Sullivan d

a Institute for Medical Technology Assessment/Institute of Health Policy & Management, Erasmus University Rotterdam, P.O. Box 1738,3000 DR Rotterdam, Netherlands b Department of Medical Oncology, University of Groningen, University Medical Center Groningen, P.O. Box 30001, 9700 RB Groningen, Netherlands c Department of Medical Oncology, Erasmus MC Cancer Institute, P.O. Box 5201,3008 AE Rotterdam, Netherlands d Kings Health Partners Cancer Centre, Institute of Cancer Policy, Bermondsey Wing, Guy's Hospital Campus, London SEI 9RT, UK

ABSTRACT

The costs of cancer care grow exponentially. It has been argued that there is a linear relation between costs and outcome: the more a country spends on cancer care, the better the outcome. We try to dispel this myth, by showing that the relation is not linear at all and by describing other factors in the cancer care delivery process that have an impact on outcome.

We show that there is a correlation between health care expenditure and life expectancy at birth, but that there is no correlation between number of deaths per 100,000 and cost per person spent on cancer in general, neither in lung, breast, colorectal and prostate cancer. Furthermore, a decrease in survival can be related to accessibility, affordability or equity issues, but also to factors such as life style. In the real world the process of cancer delivery is complex and dynamic, with many (potential) innovations. When efficacy is proven and an innovation is considered clinically relevant, the innovation has to be incorporated in evidence based clinical guidelines. However, implementation in such a guideline is still no guarantee for optimal adoption and diffusion of an innovation.

Cancer care delivery also goes beyond matters related to health-systems and cancer costs, new technologies, reimbursement agencies, hospitals, and health-care professionals by increasingly involving shared decision making. An optimal process of cancer care delivery consists of the use of new and existing diagnostic tests and treatment strategies of high quality and is effective, safe, patient centred, efficient and timely. Such health system is highly recommended and all stakeholders in society will benefit.

© 2014 The Authors. Published by Elsevier Ltd. All rights reserved.

ARTICLE INFO

Article history:

Available online 15 February 2014

Keywords:

Cancer

Outcome Adoption Diffusion Innovation Cancer drugs Dynamic process Clinical relevance Guidelines

Contents

Introduction ........................................................................................................................................................................................................................................................................................22

Adoption and diffusion of innovations in cancer care..................................................................................................................................................................................................25

Evidence based clinical guidelines..........................................................................................................................................................................................................................................26

Clinical relevance versus statistical significant outcome............................................................................................................................................................................................26

Discussion............................................................................................................................................................................................................................................................................................27

Conflict of interest............................................................................................................................................................................................................................................................................28

References............................................................................................................................................................................................................................................................................................28

* This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial-No Derivative Works License, which permits non-commercial use, distribution, and reproduction in any medium, provided the original author and source are credited.

** This paper is part of the Special Issue 'Delivering Affordable Cancer Care in High Income Countries: Papers from a Special Session of Oncology At The Limits, 13th - 15th February 2014'.

* Corresponding author. Tel.: +31 10 408 8555.

E-mail address: uyl@bmg.eur.nl (C.A. Uyl-de Groot).

Introduction

There is a widely held belief in public policy circles that higher spending results in better outcomes for cancer patients [1,2]. Yet the evidence for this is complex and many associations do not hold up on further scrutiny [3]. Health expenditures differ worldwide and generally there is a correlation between higher spending on health-care per se and composite outcome measures of health such as life expectancy. In developed countries total expenditures on health care have continued to increase substantially. Between 1995 and 2010 this rise in the USA was from 13.6% to 17.6% of its Gross Domestic Product (GDP) (Table 1) [4]. In Europe the trend is

2213-5383/$ - see front matter © 2014 The Authors. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/jjcpo.2014.01.001

Table 1

Health expenditures as % of GDP and per capita and life expectancy at birth in USA and European Union, 1995 and 2010 (cost in €).

Health expenditure, total (% ofGDP) Health expenditures percapita Life expectancy at birth

1995 2010 1995 2010 1995 2010

United States 13.6 1761 2743 6026 7562 78 54

European Union:

Austria 9.58 1098 2105 3634 7672 80 38

Belgium 7.61 1049 1564 3329 76 • 84 80 23

Bulgaria 5.23 7.58 60 351 71.05 73.51

Croatia 6.85 7.81 237 769 72.08 76.48

Czech Republic 6.69 7.46 274 1027 73.07 77.42

Denmark 8.13 11.12 2070 4577 75.21 79.1

Estonia 6.32 6.35 121 657 67.54 75.43

Finland 7.85 8.96 1471 2895 76.41 79.87

France 10.36 11.67 2007 3381 77.75 81.37

Germany 10.11 11.51 2287 3407 76.42 79.99

Greece 8.67 10.79 779 2103 77.59 80.39

Iceland 8.43 9.28 1620 2685 77.98 81.9

Ireland 6.57 9.19 906 3105 75.57 80.9

Italy 7.22 9.55 1052 2377 78.01 81.74

Latvia 5.78 6.66 84 558 66.39 73.48

Lithuania 5.37 7.05 73 572 69.01 73.27

Luxembourg 5.57 7.85 2061 5998 76.51 80.63

Malta 5.63 8.49 433 1249 77.14 80.95

Netherlands 8.33 12.11 1652 4160 77.4 80.7

Norway 9.14 9.34 2285 5884 77.74 81

Poland 5.48 6.98 146 623 71.89 76.2

Portugal 7.52 10.73 640 1688 75.31 79.03

Romania 3.22 5.95 39 335 69.46 73.46

Slovak Republic 6.06 9.01 162 1058 72.25 75.11

Slovenia 7.46 8.97 575 1511 73.96 79.42

Spain 7.44 9.6 826 2120 77.98 81.63

Sweden 7.96 9.56 1676 3447 78.74 81.45

Switzerland 9.33 10.89 3126 5636 78.42 82.25

United Kingdom 6.75 9.6 997 2558 76.84 80.4

Average European Union 8.63 10.34 1205 2441 75.8 79.63

Source: Data Worldbank [4], exchange rate: 1S-0.732D

similar, albeit less marked, indicative of important differences in the level and in the rate of growth of health care expenditures between USA and Europe. In the USA on average €6026 is spent per capita compared to the European average of €2441. The latter topped by Luxembourg and Norway with €5998 and €5884 per capita, respectively, compared to €335 and €351, respectively, for Romania and Bulgaria (Table 1).

As health-care expenditure has been increasing, so has life expectancy. In the USA the life expectancy at birth has increased from 75.6 years in 1995 to 78.5 years in 2010 (Table 1). In the 27 member states of the European Union (EU) the average life expectancy at birth improved from 75.8 years in 1995 to 79.6 years in 2010. The difference between EU countries with the highest and lowest life expectancies is around 8 years for women and 12 years for men [6]. Factors influencing this gain were rising living standards, improved lifestyle, better education and improved access to health services [6]. Fig. 1 shows the correlation between the health

90 85 80

Years 75

70 65 60

• EU countries and USA

0 5 10 15

Health expenditures, total (%GDP)

Fig. 1. Relation life expectancy at birth and health expenditure (%GDP). Source: Data Worldbank [4], Data derived from Table 1.

care expenditure and life expectancy (Pearson correlation: 0.52, p = 0.0032).

But what is the situation for cancer? Cancer is the second leading cause of mortality in most developed countries after cardiovascular diseases. In 2008, 2.45 million people were diagnosed with cancer in the EU. While changes in lifestyle and advances in prevention, early detection, and treatment have considerably reduced cancer incidence and mortality, 1.23 million people still died in the EU of this disease [3]. Death rates for all types of cancer in males and females have modestly declined since 1995, although the decline has been less than for cardiovascular diseases [6]. The economic burden of cancer is high. In the EU in 2009 the total costs were €126 billion, €51.0 billion of which related to EU health-care systems. Taken by EU country, the economic cost varied from €16 per person in Bulgaria to €184 per person in Luxembourg [3]. But is there a direct causal relationship between expenditure on cancer care and outcomes?

It has been argued that there is a linear relation between costs and outcomes in cancer care: higher spending results in better outcomes [1,7]. Philipson et al. report that the USA spends more on cancer care per capita than any European country, but this is paralleled by greater survival gains. The adjusted average survival from diagnosis was 11.1 year for cancer patients diagnosed in the US from 1995-1999, compared with 9.3 years among the European countries [1]. The authors suggest the value of these additional survival gains has exceeded the additional costs of care. However, there are several reasons why survival is not a good indicator to compare countries.

De Vries et al. already described that a decrease in survival apart from deterioration of care or of access to it, can also be attributed to improved diagnosis of premalignant lesions, to deleterious changes in the distribution of prognostic factors and changes in the

Table 2

Death rates per 100.000 and health care costs of all cancers and of lung, breast, colorectal and prostate cancer in European countries in 2009, by country.3

Death rates per 100,000 [5]

Adjusted cost per person (€) [3]

All Lung Breast Colorectal Prostate All Lung Breast Colorectal Prostate

cancers cancer cancer cancer cancer cancers cancer cancer cancer cancer

Austria 206.30 37.8 28.8 22.3 33.3 119 11 16 13 12

Belgium 216.40 52.3 34.2 24.2 29.3 71 6 9 9 8

Czech Republic 253.10 47.9 25.5 35.9 35.0 104 9 13 13 11

Denmark 243.20 58.4 35.5 31.9 47.9 69 6 8 8 8

Estonia 237.40 40.4 26.1 28.2 55.7 82 7 13 11 7

Finland 176.90 32.2 23.5 17.9 34.8 127 10 16 13 14

France 211.90 42.5 28.8 22.9 31.6 97 6 13 9 13

Germany 207.00 40.5 29.5 24.8 30.7 171 15 27 20 20

Greece 203.30 47.2 27.1 17.6 26.8 128 11 20 10 16

Ireland 242.50 50.6 34.9 27.7 41.7 88 8 9 10 7

Italy 210.60 42.4 28.2 23.1 24.0 96 8 9 11 8

Luxembourg 215.20 46.9 31.4 27.3 37.5 141 16 20 17 14

Netherlands 239.30 57.0 32.9 28.5 39.1 123 12 18 16 8

Poland 251.8 58.7 24.2 28.9 33.4 78 11 9 9 5

Portugal 201.30 28.5 24.3 29.8 36.9 61 4 8 6 7

Slovak Republic 246.10 41.4 26.2 37.8 33.9 103 9 14 11 10

Slovenia 258.20 45. 33.2 35.8 52.6 90 7 10 9 10

Spain 197.80 40.8 21.6 27.3 26.5 96 5 12 10 11

Sweden 192.80 31.4 23.1 23.3 51.1 92 7 10 6 11

United Kingdom 228.50 51.3 31.2 23.1 37.1 92 8 10 10 7

Source: OECD statistics [5], Fuengo-Fernandez [3]. a Selection of countries based on available data.

distribution of socio demographic characteristics in a population [8]. Furthermore, differences in survival may be influenced by differences in adherence to a healthy life style after cancer diagnosis as healthy lifestyle is increasingly considered to improve outcomes [9]. Moreover the number of patients with a second primary tumour is increasing [10,11]. So there is a shift in the composition of the patient population. Since treatment of second and later primary tumours, due to limitations related to successful previous treatments (chemotherapy maximum dose already achieved; drug-resistance issues), can be hindered this may negatively affect overall survival figures of these tumours.

In order to have a better comparison between cost and outcome, we studied the relation between cancer mortality and costs in EU countries (Table 2). Fig. 2 shows that there is almost no correlation between number of deaths per 100,000 and cost spent on cancer per person (Pearson correlation: -0.31, p = 0.1836)), implying that countries reporting low number of deaths per 100,000 did not spend more money per person.

Considering this relation for lung, breast, colorectal and prostate cancer the results were even more pronounced. For lung cancer the correlation was positive, implying that on average less money spent on lung cancer did not result in worse outcome (Pearson correlation: 0.19, p = 0.4258)) (Fig. 3.1). For breast, colorectal and prostate cancerthere also was no linear relation, the correlations were -0.03

300 250 200

Deaths per 100,000 150

♦ EU countries

50 100 150

Adjusted costs per person (in €)

Fig. 2. Relation cancer mortality versus costs in EU countries (2009). Source: OECD statistics [5], Fuengo-Fernandez [3], 'Selection of countries based on available data, Data derived from Table 2.

(p = 0.8915), -0.07 (p = 0.7688) and -0.30 (p = 0.1913), respectively (Figs. 3.2-3.4).

With respect to costs of cancer care new cancer medicines dominate the public debate [12]. However, although costs of drugs are a significant factor accounting for around 27% of total costs, hospital inpatient care is responsible for over 56% of direct cancer care costs, a factor that is often overlooked [3]. Moreover a disproportional amount of the money is spent during the last 30 days of life, particularly involving inpatient care. In the USA the costs of the latter period were almost 30% of total cancer care costs and even in the EU this was as high as 20% [13,14]. Therefore it is important to discuss drug expenditure in cancer as part of a much wider issue, that encompasses the structural, organisational and cultural aspects of cancer care.

The benefits of novel expensive anticancer drugs with respect to survival are increasing. While major impacts have been observed with some agents in a limited number of tumour types, this was mainly in the non-curative, advanced setting where most of those and other targeted agents were studied. Assuming a higher likelihood of benefit and success, the next generation of molecularly targeted medicines as well as immunotherapies are increasingly tested in the adjuvant and neo-adjuvant settings.

Given these developments clinicians as well as lay press emphasise the need for more balanced pricing of cancer drugs. In economic terms, price must reflect value and/or society's willingness to pay, a subjective position that harnesses many public, political and fiscal factors [15]. However, the last two decades prices of newly introduced cancer drugs have increased more than twentyfold, from an average of €300-500 per patient per month (carboplatin) in the early nineties to more than €10,000 per patient per month (ipili-mumab). In 2012,12 cancer drugs were approved by the FDA, 11 of which were priced above €73,000 per treatment per patient [16].

In this respect a uniform system of expressing costs per value is missing. Cost-effectiveness analysis is widely used to support decision making for which costs per QALY is frequently used. When prices are based on real value, it could be expected that the cost per QALY would largely remain the same. Those treatments with the best balanced cost per QALY gained would then be reimbursed. The UK National Institute for Health and Care Excellence (NICE) for instance, states that for drugs the cost per QALY should not

70 60 50

Deaths per 40 100,000 30

70 60 50

Deaths per 40 100,000 30

20 10 0

70 60 50

Deaths per 40 100,000 30

♦ EU countries

Adjusted costs per person (in €)

^—ti ♦

♦ EU countries

10 20 Adjusted costs per person (in €)

70 60 50

Deaths per 40 100,000 30

20 10 0

♦ EU countries

Adjusted costs per person (in €)

♦ EU countries

Adjusted costs per person (in €)

Fig. 3. (1) Mortality versus costs in EU countries: lung cancer (2009), (2) Mortality versus costs in EU countries: breast cancer (2009), (3) Mortality versus costs in EU countries: colorectal cancer (2009), (4) Mortality versus costs in EU countries: prostate cancer (2009).

Source: OECD statistics [5], Fuengo-Fernandez [3], 'Selection of countries based on available data, Data derived from Table 2.

normally exceed €36,180 (1 £^'"€1.206". However using the cost per QALY as a principle denominator is not accepted by everyone, since not all relevant factors, such as conflicting outcomes, severity of disease, rarity of disease, the limited availability of other treatment options and budget impact, are taken into account in this parameter. Moreover cancer medicines are often used for different cancers and/or within one cancer at different stages of disease (e.g. trastuzumab), with different cost per QALY outcomes. To address this technology assessment programmes are increasingly looking to use of multi-criteria decision analysis (MCDA) as an approach to prioritise expensive medicines. However, practical applications in cancer care are still rather limited.

Beyond costs there are clearly wider issues that affect outcome. There is no easy direct, linear relationship between spending and outcomes, such as mortality, survival or progression free survival, in cancer care (e.g. see Figs. 2 and 3). In the real world the process of cancer care delivery is complex and depends on several other factors like the health-care system, the adoption and diffusion of the innovations in a system, quality of the clinical guidelines, the (economic) behaviour of several stakeholders and characteristics of patients. These other factors are however much more difficult to tackle then, for example the cost of cancer medicines and are often neglected. We need to aim for health-care delivery systems in which cancer care is of highest possible quality, effective, safe, personalised, timely, efficient and equitable [17]. But how is this to be achieved?

Adoption and diffusion of innovations in cancer care

When an innovation, e.g. a new diagnostic test and/or treatment opportunity, is developed the diffusion of the innovation depends on several factors. Diffusion shows how an innovation spreads among (potential) users. According to Rogers, diffusion of an innovation is a five-step process percolating through a series of communication channels over a period of time among the members

of a similar social system [18]. Factors of importance that determine diffusion are knowledge, awareness, persuasion and interest. Cancer care is rich in evidence-based innovations. However, the adoption of these innovations is quite heterogeneous and differs between countries and even between regions within countries. The way a new innovation is perceived, the characteristics of the individuals who may adopt the change and contextual and managerial factors within the care delivery determine to a great extent the rate of diffusion [19]. In this respect important stakeholders are the reimbursement agencies, health insurers, hospitals, doctors and patients. However, when an innovation is reimbursed by a reimbursement agency, access to the innovation could still be different between hospitals or regions within one country [20]. The adoption of an innovation often follows an S shape, when plotted over time. Fig. 4 shows that within one health system the diffusion of an innovation, i.e. administration of bortezomib in relapsed/refractory multiple myeloma (registered in 2004 for 2nd and 3rd line therapy based on a phase II trial [21]; phase III data became available in

Proportion of multiple 15 myeloma patients

♦ Average proportion Netherlands ■ Minimum

2003 2004 2005 2006 2007 2008 2009 2010 Year

Fig. 4. Uptake of bortezomib in the Netherlands, period 2004-2010. Source: Blommestein [20].

2005) [22], indeed followed an S shape, but the rate of adoption differed remarkably between the regions [20].

It is well known that doctors who have participated in clinical trials are more eager to adopt an innovation than colleagues who did not. The 'trial' doctors and 'trial' centres already have experience with the innovation, thus when there is a proven clinical benefit they are inclined to administer the innovation to their patients. These are the so called early adopters. In this respect the role of key clinical leaders is very important as they have most influence during the adoption process. Key opinion leaders also have more contacts with developers of innovations, have greater exposure to the mass media and are involved in the development of clinical guidelines. Whilst rapid diffusion of new innovations has been championed by a number of parties, mostly those constituents focused on medical technologies, the drive to bring 'innovations' to clinical practise is not necessarily a win-win situation [19]. Key opinion leaders can be too enthusiastic about an innovation. Sometimes there is too much focus on the gain in progression free survival, rather than in overall survival, side-effects or even cost. For example in 2011 the Food and Drug Administration (FDA) decided to revoke an approval of 2008 for the breast cancer drug bevacizumab. The FDA concluded that "based on new information the considerable risks of taking bevacizumab cannot be justified when there is no proof of a benefit in breast cancer patients" [23]. While the drug is still available for other types of cancer treatment, the approval for use in breast cancer was retracted. Furthermore whilst innovation covers all forms of technology and practice the reality is that apart from cancer medicines and associated diagnostics there has not been a similar drive in other areas, such as mental health innovations in cancer [24], or even in the major modality of cancer control, surgery, to deliver in these areas.

Evidence based clinical guidelines

Even where cost effective pathways and/or technological innovations have been demonstrated the subsequent diffusion into health-care systems is dependent on clinical take up. This demands inclusion in cancer clinical guidelines. However, there are reservations about the content and reliability of oncology clinical practice guidelines [25]. Many recent published guidelines did not meet the criteria in the checklist and many existing guidelines are outdated. Reames et al. concluded that there is much room for improvement to make guidelines as methodologically sound and evidence based as possible [25]. In addition these guidelines are not very easy to use in daily practice. Moreover especially in case of drug trials there are in general restricting eligibility criteria, which means that patients with significant co-morbidity or another primary tumour in the past are often excluded from participation. Guidelines do not frequently take into account the limitations of clinical trials in reflecting the real world. This is one of the reasons why futile treatment may occur. In this respect it is of interest that in 2012 the American Society Clinical Oncology (ASCO) recommended in their "Choosing Wisely items" [26] to eliminate treatment that is unlikely to be effective in patients who meet all of the following criteria: Patients with low performance status (3 or 4); who have not benefited from prior standard therapy; who have no further standard treatment options for their disease; and who are not eligible for a clinical trial. For these patients, emphasis should be placed on palliative and supportive care, that can lead to evidence-based increase quality of life and, in some cases, increase survival [27].

Cancer care is a very dynamic process; the number of potential innovations is high. For example at the moment there are over 900 new molecular entities in (drug-)development as well as vast numbers of new approaches in screening, radiotherapy, pathology and surgery. The cancer research space is complex and very dynamic. Developing and keeping up to date cancer care guidelines is a labour

intensive and time-consuming process that needs input from many stakeholders, since delivering cancer care is typically multidisci-plinary. Ensuring health-care systems follow appropriate cancer clinical guidelines is an essential step in delivering better outcomes. However, this first requires appropriate and timely guidelines to be constructed.

Clinical relevance versus statistical significant outcome

A registered drug is not necessarily a relevant drug and a statistical significant difference is not the same as a clinical meaningful benefit. At the moment, several organisations, such as ASCO, NICE, European Society of Medical Oncology and the Dutch Committee on Expensive Cancer Drugs, are dealing with this question. Depending on several factors such as cancer type, line of therapy and subsets of patient groups different criteria are formulated according to clinical measurable outcomes, like response, progression free survival, overall survival, adverse events and quality of life. In this respect, especially treatments used in the curative setting (primary endpoint: overall survival) and treatments used in the non-curative setting (any other endpoint) should be assessed differently.

There is of course a risk-benefit trade off between improvement in (overall) survival on one hand and toxicity and quality of life issues on the other, that should be taken into account. Ellis et al. already drafted ASCO recommendations in order to stimulate clinical trials with new drugs or regimens based on molecular or biologic markers which will have the potential to significantly improve the lives of patients (in terms of survival and quality of life) and lead to real advances in therapy [27]. The result will be "smaller and smarter" and more efficient trials [27,28]. Improvements of at least 5-20% in overall survival are seen as clinical meaningful outcomes [27,29].

As previously discussed, the adoption of an innovation is not necessarily implemented when it is first reported in a clinical guideline. However, there are also other factors why doctors did not use the innovation. Sometimes they are not aware and need information (feedback) about the treatments and outcomes of treatments they have given to patients and compare these with other colleagues and/or hospitals. This information must be evaluated and used to define a plan to improve their functioning. As such medical professional organisations are performing audits and increasingly clinical, process and outcome data are collected. Patient registries are useful tools [30,31]. Key elements are collecting data, sharing data and communication between doctors [32]. Although adoption is rather an individual process, it is advisable to give the feedback on a hospital or department level rather than on an individual level [32]. In this respect, ASCO's Quality Oncology Practice Initiative (QOPI®) programme is also of interest [33]. This initiative is an oncologist-led, practice-based quality assessment and improvement programme. The goal is to promote excellence in cancer care by helping practices create a culture of self-examination and improvement. The process employed for improving cancer care includes measurement, feedback and improvement tools for haematology-oncology practices [33].

Cancer care delivery also goes beyond matters related to health-systems and cancer costs, new technologies, reimbursement agencies, hospitals, and health-care professionals. The position of patients is also becoming central to the debate on what sort of cancer care systems society expects. Whilst public policy may rest on macro economics in cancer care the reality is that in many countries the popular view of what is expected for patients shapes much of the strategic discourse. However, to implement shared decision making (SDM) in daily practice there is a need to significantly change the philosophy, culture and roles of patients and doctors' [17].

Fig. 5. Process of cancer care delivery.

Discussion

We here challenge the linear relation between costs and outcome in cancer care. Such an assertion is one of the central myths

of cancer today. The fact this is not the case calls for more insightful and broad learning from best practices across health-care systems. This is of major importance as the costs of cancer care delivery are increasing in an unsustainable manner. Increasing health care

expending will result in limited and unequal access, particularly with regard to new innovations with high acquisition costs, such as new oncology drugs, radiotherapeutic advances, laboratory tests etc. Policy makers have to make reimbursement decisions considering both rapid and equal accessibility to promising technologies as well as decisions about how to divide available resources among diseases.

It has been noted that new cancer drugs often reach USA patients earlier than European patients [1]. To solve the problem of delay in access and frequently limited available evidence the 'coverage with evidence development' policy (or "conditional approval") is an option: new drugs are reimbursed under the condition that additional research will be conducted [30,31]. However, we describe that there are also other factors involved in the cancer care delivery process suggesting that limitations in availability of new technologies may be beneficial if there is limited knowledge of the technologies clinical impact and cost effectiveness. Fast does not always mean a better deal for patients or society, despite the prevailing public policy which is for ever quicker approval.

There are many clinical practice guidelines in cancer care and these are known to have a substantial effect on practice patterns and patient outcomes [25]. However, they have substantial deficiencies. Furthermore, due to changes in the population, the dynamic and complex process of cancer care, including the rapid growth of new innovations, the applicability of the guidelines in daily practice is rather difficult and also limited. This is pointed out in Fig. 5, which presents the dynamic process of delivering cancer care.

One reason is that, due to an ageing population, there will be more elderly patients with cancer, involving more patients with comorbidity. In these situations, doctors have to balance between a longer life for patients and the side effects of treatment, but without having much evidence available for this patient group. There is an increasing need for optimal treatment of cancer patients with a variety of 'concomitant' chronic conditions [34]. As the evidence is mostly not found in clinical trials, it should be collected from daily practice, e.g. by patterns of care studies.

SDM between doctors and patients is currently receiving much attention. It is assumed that the involvement of patient preferences favours less extensive treatment. Therefore, SDM has also been promoted as a strategy to reduce overtreatment and costs [35,36]. However, SDM could also lead to higher and sometimes unrealistic demands, leading to overtreatment. Recently it was already indicated that the increasing expectations about the role of SDM in clinical and health policy in general and in cancer in particular should not be overestimated [37,38].

In the end, appropriate use of innovations and accessibility depend on various factors, such as regulatory and healthcare system, organisational characteristics, (dynamic) process of innovations, and socio-cultural characteristics of numerous stakeholders (including patients). An evaluation and comparison of health policies around cancer across Europe will provide more insight in this dynamic and complex process. Furthermore, it will give valuable information that can enhance evidence-based decision making for both health care providers and policy makers. Such an evaluation could improve appropriate use of new and existing diagnostic tests, treatment strategies and pathways in order to reach a cancer care delivery system of optimal quality, effective, safe, person centred and timely. When this system is also efficient and equitable all stakeholders in the society will benefit from it.

Conflict of interest

There is no conflict of interest

References

Philipson T, Eber M, Lakdawalla DN, Corral M, Conti R, Goldman P. An analysis of whether higher health care spending in the United States versus Europe is 'Worth It' in the case of cancer. Health Aff 2012:31:4667-75. Sun E, Jena A, Lakdawalla D, Reyes C, Philipson T, Goldman D. An economic evaluation ofthe war on cancer. Cambridge, MA: National Bureau of Economic Research: 2009. Contract No: 15574.

Luengo-Fernandez R, Leal J, Gray A, Sullivan R. Economic burden of cancer across the European Union: a population-based cost analysis. Lancet Oncol 2013:14:1165-74.

http://data.worldbank.org/news/world-development-indicators-2012-now-available [accessed 05.01.14]. http://stats.oecd.org/ [accessed 05.01.14].

Health at a glance. Europe: OECD Publishing: 2012 http://dx.doi. org/10.1787/9789264183896-en [accessed 05.01.14].

Porter ME. A Strategy for health care reform - toward a value-based system. N Engl J Med 2009:361:109-12.

De Vries E, Karim-Kos HE, Janssen-Heijnen MLG, Soerjomataram I, Kiemeney LA, Coebergh JW. Explanations for worsening cancer survival. Nat Rev Clin Oncol 2010:7:60-3.

Tao L, Wang R, Gao YT, Yuan JM. Impact of postdiagnosis smoking on long-term survival of cancer patients: the Shanghai cohort study. Cancer Epidemiol Biomarkers Prev 2013:22:2404-11.

Rheingold SR, Neugut AI, Meadows AT, Rheingold SR, Neugut AI, Uldrick T, Meadows AT. Treatment-related secondary cancers. In: Hong WK, Bast RC, Hait WN, Kufe DW, Pollack RE, Weichselbaum RR, Holland JF, Frei EF, editors. Holland-Frei cancermedicine. 8thed. Shelton: People's Medical Publishing House: 2009. p. 1915-20.

Oeffinger KC, Baxi SS, Novetsky Friedman D, Moskowitz CS. Solid tumor second primary neoplasms: who is at risk, what can we do? Semin Oncol 2013:40:676-89.

Lewison G, Tootell S, Roe P, Sullivan R. How do the media report cancer research? A study ofthe UK's BBC website. Br J Cancer 2008:99:569-76. Raphael C, Ahrens J, Fowler N. Financing end-of-life care in the USA. J RSoc Med 2001:94:458-61.

Silvia Francisci S, Stefano Guzzinati S, Maura Mezzetti M, Emanuele C, Francesco G, Guido M, et al. Cost profiles of colorectal cancer patients in Italy based on individual patterns of care. BMC Cancer 2013:13:329.

Pfister DG. The just price of cancer drugs and the growing cost of cancer care: oncologists need to be part ofthe solution. J Clin Oncol 2013:31:3487-9. Experts in Chronic Myeloid Leukemia The price of drugs for chronic myeloid leukemia (CML) is a reflection of the unsustainable prices of cancer drugs: from the perspective of a large group of CML experts. Blood 2013:121: 4439-42.

The Health Foundation. Summit report: leading the way to shared decision making: 2012 http://www.health.org.uk/public/cms/75/76/313/ 3293/Leading%20the%20way%20to%20shared%20decision%20making. pdf?realName=RmwIPj.pdf [accessed 05.01.14]. Rogers EM. Diffusion of innovations. New York: The Free Press: 1962. Berwick DM. Disseminating innovation in health care. JAMA 2003:289:1969-75.

Blommestein HM, Franken MG, Verelst SGR, van Agthoven M, Huijgens PC, Uyl-de Groot CA. Access to expensive cancer drugs in Dutch daily practice: should we be concerned? NethJ Med 2014 [in press].

Richardson PG, Barlogie B, Berenson J, Singhal S, Jagannath S, Irwin D, et al. A phase 2 study of bortezomib in relapsed, refractory myeloma. N Engl J Med 2003:348:2609-17.

Richardson PG, Sonneveld P, Schuster MW, Irwin D, Stadtmauer EA, Facon T, et al. Bortezomib or high-dose dexamethasone for relapsed multiple myeloma. N Engl J Med 2005:352(24):2487-98.

http://www.fda.gov/NewsEvents/Newsroom/ucm280585.htm [accessed 10.01.14].

Purushotham A, Bains S, Lewison G, Szmukler G, Sullivan R. Cancer and mental health - a clinical and research unmet need. Ann Oncol 2013:24:2274-8. Reames BN, Krell RW, Ponto SN, Wong SL. Critical evaluation ofoncology clinical practice guidelines. J Clin Oncol 2013:31:2563-8.

http://www.choosingwisely.org/doctor-patient-lists/american-society-of-clinical-oncology/

Ellis LM, Bernstein D, Voest E, Berlin J, Sargent D, Cortazar P, et al. Defining clinically meaningful outcomes: ASCO Recommendations to raise the bar for clinical trials. Am Soc Clin Oncol 2014. http://www.asco. org/sites/www.asco.org/files/asco_meaningful_outcomes_drafLfor_website_-_05_13_13_0.pdf

Sobrero A, Bruzzi P. Incremental advance or seismic shift? The need to raise the bar of efficacy for drug approval. J Clin Oncol 2009:27:5868-73. NVMO. Systematische beoordeling van oncologische middelen. Medische Oncologie 2012:1:26-9 [In Dutch].

Franken MG, van Gils CW, Gaultney JG, Delwel GO, Goettsch W, Huijgens PC, et al. Practical feasibility of outcomes research in oncology: lessons learned in assessing drug use and cost-effectiveness in The Netherlands. Eur J Cancer 2013:49:8-16.

[31] van Gils CW, de Groot S, Redekop WK, Koopman M, Punt CJ, Uyl-de Groot CA. Real-world cost-effectiveness of oxaliplatin in stage iii colon cancer: a synthesis of clinical trial and daily practice evidence. Pharmacoeconomics 2013:31:703-18.

[32] Ilgen DR, Hollenbeck JR, Johnson M, Jundt D. Teams in organizations: from input-process-output models to IMOI models. Rev Psychol 2005;56:517-43.

[33] http://qopi.asco.org/References.Htm [accessed 10.01.14].

[34] Louwman WJ, Aarts MJ, Houterman S, van Lenthe FJ, Coebergh JW, Janssen-Heijnen ML. A 50% higher prevalence of life-shortening chronic conditions among cancer patients with low socioeconomic status. Br J Cancer 2010;103:1742-8.

[35] Oshima E, Emanuel EJ. Shared decision making to improve care and reduce costs. N Engl J Med 2013;368:6-8.

[36] Wennberg JE, O'Connor AM, Collins ED, Weinstein JN. Extending the P4P agenda. Part 1. How Medicare can improve patient decision making and reduce unnecessary care. Health Aff (Millwood) 2007;26:1564-74.

[37] Katz SJ, Hawley S. The value of sharing treatment decision making with patients: expecting too much? JAMA 2013;310:1559-60.

[38] Helwick C. Sharing treatment decision-making with patients: where's the evidence ofvalue. JAMA 2013;310:1559-60.