Health-Care Accessibility

The term health-care accessibility originally was used narrowly to describe the ability of patients to obtain entry into the health system. Today, it is used more broadly to encompass all of the factors that influence the level of use of a health service in relation to the level of need for the service in a population. The concept of access has been described as representing the overall “degree of fit between the clients and the system.” Several factors have been shown to influence the overall degree of fit. Availability describes the volume and type of services available in relation to the number of patients and their needs. Spatial accessibility describes the relationship between the location of supply of service and the location of the patients who need the service, taking into account travel times and costs. Accommodation describes the extent to which the system is designed to facilitate patient access to service, for example, by operating at convenient hours or by providing lodging for those who require treatment that is only available far from their homes. Affordability describes the relationship between prices and the ability of patients to pay. It also encompasses indirect costs, for example, loss of earnings during treatment that may deter use of the service. Awareness describes the extent to which those who need the service are aware that it is available and that they might benefit from it.

Quality in Health Care

Almost half a century ago, Donabedian defined the quality of health care as “a property of, and a judgement upon, some definable unit of health care, and that care is divisible into at least two parts: technical and interpersonal.” The quality of technical care is measured by the extent to which “the application of medical science and technology maximizes its health benefits without correspondingly increasing its risks.” The quality of interpersonal care is measured by “how well the physician–patient interaction meets the socially defined norms of the relationship.” Today, the Institute of Medicine describes health-care quality as “the degree to which health services for individuals and populations increase the likelihood of desired health outcomes and are consistent with current professional knowledge.”

Donabedian also provided a framework for evaluating the quality of health care in terms of structure, process, and outcome. The term structure is defined broadly to include facilities, equipment, personnel, and organizational structures. The term process includes both the type of care that is delivered and the way in which it is delivered. The term outcome refers to the consequences of the care that has been provided. Donabedian reasoned that optimal process is necessary for optimal outcome; that adequate structure is necessary, although not sufficient, for optimal process; and that outcomes may be enhanced by identifying and correcting deficiencies in structure and/or process.

Efficiency in Health Care

The resources available for health services are always limited, even in high-income nations. What is achievable in terms of cancer control depends on the total health-care budget, on how much of that total budget is directed to cancer care, and on how efficiently the available resources are used in providing cancer care. Efficiency measures whether we are getting the best value for money from the available health-care resources. Inefficiency is said to be present when using those resources in a different way would provide greater health benefits. Health economists distinguish between technical efficiency, productive efficiency, and allocative efficiency. As stated by Palmer and Torgerson, “technical efficiency addresses the issue of using given resources to maximum advantage; productive efficiency of choosing different combinations of resources to achieve the maximum health benefit for a given cost; and allocative efficiency of achieving the right mixture of healthcare programs to maximize the health of society.” Methods for measuring each of these quantities are well established and have been used to address some important issues in the treatment of lung cancer, but the sphere of health economics is beyond the scope of this chapter and therefore we will not deal with it in detail here.

In this chapter, we will review the results of research studies that have sought to optimize the accessibility and quality of programs of care for patients with lung cancer. This work involves identifying barriers to optimal care as well as designing and evaluating interventions to overcome those barriers. However, before one can identify deviations from optimal health system performance, one must first identify appropriate indicators of performance and establish standards of performance with respect to those indicators. We will therefore begin by reviewing the prescriptive research that has been undertaken to establish standards of care for patients with lung cancer. We will distinguish between standards of care for the individual patient and standards for the operation of the health programs that are required to deliver care to a population of patients. Standards of care for the individual patient should be based, whenever possible, on the results of randomized clinical trials that directly compare the outcomes of alternative forms of treatment. Likewise, standards for the operation of health programs should be based on the results of randomized trials that directly compare the effectiveness of alternative approaches to health-care delivery or at least on the results of well-controlled observational studies. However, we will show that the empirical evidence that underpins program standards today is usually much weaker than the evidence that supports guidelines for the care of the individual patient.

Guidelines for the Care of Individual Patients With Lung Cancer

Guidelines for the treatment of lung cancer have been developed by many different agencies and organizations around the world, including specialty societies such as the American Society of Clinical Oncology (ASCO), the European Society for Medical Oncology, the American College of Chest Physicians, and the American Society for Radiation Oncology; groups of health-care institutions such as the National Comprehensive Cancer Network (NCCN) in the United States; and government agencies such as the National Institute for Health and Care Excellence in the United Kingdom, Cancer Care Ontario in Canada, and the Cancer Council Australia. Some of these agencies attempt to provide comprehensive guidelines that cover the entire spectrum of possible presentations of the disease, whereas others have focused on providing detailed guidelines for management in specific clinical contexts. Most guidelines are written primarily for the use of physicians, but some agencies also provide a version that targets patients directly. It is recognized that practice guidelines that are created in one country may not be applicable in other societies with very different resources and/or different populations of patients. The NCCN, an alliance of 23 major cancer centers in the United States, is now endeavoring to provide international adaptations and translations of its guidelines to make them suitable for use in countries with different levels of economic development. Nonetheless, it is probably preferable for each society to develop its own guidelines. Repeating work that has already been done elsewhere may seem to be a waste of resources, but the process of guideline development is important in its own right and may have as much normative effect on practice as the guideline itself.

Challenges to the Development and Application of Practice Guidelines

Variation in patient values and the biologic heterogeneity of tumors pose particular challenges for the creation and application of treatment guidelines.

Do Guidelines Guide Practice?

A quarter of a century ago, Lomas et al. famously asked this question after the dissemination of a nationally endorsed consensus statement recommending decreases in the use of cesarean deliveries in Canada. Their answer was a qualified “no.” Lomas et al. found that, although one-third of the hospitals and obstetricians reported changing their practice as a consequence of the guidelines, actual practice changed very little. The authors concluded that “guidelines for practice may predispose physicians to consider changing their behavior, but unless there are other incentives or the removal of disincentives, guidelines are unlikely to affect rapid change in actual practice.” It is therefore important to evaluate the extent to which guidelines for the treatment of lung cancer actually guide practice.

A number of studies have evaluated the degree of concordance between clinical practice and guidelines for the treatment of lung cancer. Some studies have assessed whether treatment recommendations were concordant with guidelines, whereas others have assessed whether actual treatment was concordant with guidelines.

In two recent studies, clinician recommendations for treatment were compared with the corresponding guidelines. Vinod et al. evaluated the degree of concordance between treatment recommendations and Australian practice guidelines in a study involving a cohort of patients with lung cancer who were discussed in a multidisciplinary management meeting (MDM). The MDM recommendations were deemed to be concordant if the general plan of treatment that was recommended corresponded with the guidelines. The rate of concordance was 71% for overall management, 58% for surgery, 88% for radiotherapy, and 71% for chemotherapy. Couraud et al. asked oncologists and pulmonologists specializing in thoracic oncology for their treatment recommendations in four hypothetical clinical scenarios and compared their recommendations with the corresponding French guidelines. Their criteria for concordance required consideration of the details of treatment, including the specific chemotherapeutic agents and the number of cycles of chemotherapy recommended. On the basis of these fairly strict criteria, the rate of concordance with the guidelines ranged from 25% to 63% in the four scenarios. Clinicians who worked in public practice were more likely to comply with the guidelines than those who worked in private practice. Overall, only 15% of clinicians applied the guidelines appropriately to all four cases, and 10% did not apply them in any of the cases. Not surprisingly, these two studies illustrate that the observed degree of concordance depends on how strictly concordance is defined.

A number of population-based studies have assessed the concordance between the treatment that patients actually receive and the prevailing guidelines for the treatment of lung cancer. In a population-based study in the United States that was initiated in 1996, Potosky et al. described the treatment of NSCLC and compared the observed treatment with best practice as defined by the authors based on the evidence at the time. Overall, 52% of the patients received guideline-recommended treatment, but this rate varied significantly from 41% for patients with stage IV disease to 69% for patients with stage I and II disease ( p < 0.05). The rate of guideline-recommended treatment was significantly lower for older patients, for patients who were single, and for the nonwhite population ( p < 0.05). A decade later, in a similar US population-based study, the actual treatment of NSCLC was compared with recommended treatment in the prevailing NCCN guidelines. The rate of guideline-recommended treatment was only 42% overall, 37% for patients with stage I or II disease, 58% for those with stage III disease, and 29% for those with stage IV disease. Older patients and African-American patients were less likely to receive guideline-recommended treatment. In a population-based study from the Netherlands, de Rijke et al. found that only 44% of patients with stage I–III NSCLC received guideline-recommended treatment. The rate of guideline-recommended treatment again varied according to stage and was reported to be 82% for patients with stage I or II disease, 48% for patients with stage IIIA disease, and 54% for patients with stage IIIB disease. Rates of guideline-recommended treatment were significantly lower among older patients, and higher levels of comorbidity and lower Eastern Cooperative Oncology Group (ECOG) performance status were associated with a lower rate of guideline-recommended treatment for patients with stage I and II disease. Duggan et al. reported very similar findings in a population-based study in which the actual treatment of lung cancer was compared with treatment recommended in Australian guidelines. The rate of guideline-recommended treatment was 54% for patients with small cell lung cancer (SCLC) and 51% for patients with NSCLC. Increasing age and poorer ECOG performance status were associated with lower rates of guideline-recommended treatment.

Thus, population-based studies from three continents have shown that, at best, only half of all patients with lung cancer are treated according to guidelines. All of these studies showed that older patients are less likely to receive guideline-recommended treatment. Although these studies did not consistently evaluate other patient characteristics, taken together, they provide evidence that patients with poorer performance status or higher levels of comorbidity are less likely to receive guideline-recommended treatment. These general findings have been supported by the results of several similar studies in more selected populations of patients with lung cancer. Other studies of compliance with guidelines focusing on specific clinical situations have provided similar results. Allen et al. found low rates of adherence to surgical guidelines for the treatment of operable NSCLC; Salloum et al. noted low rates of guideline-recommended treatment with respect to the use of chemotherapy in patients with stage II to stage IV NSCLC; and Langer et al. found low rates of guideline-recommended treatment with respect to the use of concurrent chemotherapy and radiotherapy in patients with limited-stage SCLC and stage I to stage III NSCLC.

Two recent studies showed that there may be cogent reasons why patients do not receive guideline-recommended treatment. Landrum et al. found that many patients who did not have surgery for the treatment of stage I or II NSCLC, as indicated in the guidelines, were either in poor health (61%) or had declined the operation (26%). Poor health was defined as encompassing advanced age, comorbidity, poor performance status, and poor lung function. In a presentation focusing on a cohort of patients with lung cancer who were discussed at an MDM, Boxer et al. noted that the main reasons why patients did not receive guideline-recommended treatment were a decline in performance status (24%), large tumor volume precluding radical radiotherapy (17%), comorbidities (14%), and patient preference (13%).

Thus, discordance between practice and guidelines for cancer treatment does not always indicate inappropriate patient care. It may instead indicate that the existing guidelines do not adequately take into account variations in the health status of patients with cancer in the general population. This consideration is particularly important in the context of lung cancer, with more than 50% of patients having at least one other substantial medical problem that may affect their care.

Does Adherence to Guidelines Improve Outcomes in the General Population?

It is important to ask whether adherence to evidence-based guidelines actually yields the improvements in outcome in the general population that would be expected on the basis of the results of the relevant clinical trials. Only a small proportion of all cases of lung cancer are included in clinical trials, and these cases are not necessarily representative of the overall population of patients with lung cancer. Furthermore, the institutions and physicians likely to engage in clinical trials are unlikely to be representative of the heath-care system as a whole.

Two studies showed that patients with lung cancer who are seen in routine practice are indeed different from those who are enrolled in clinical trials. De Ruysscher et al. found that 59% of patients with stage III NSCLC and limited-stage SCLC who were seen at their clinic would not have been eligible for entry into the clinical trials that showed a benefit from concurrent chemoradiation therapy, mainly because of their advanced age. Firat et al. found that 33% of the patients who were treated with chemoradiotherapy at their center did not meet eligibility criteria for any Radiation Therapy Oncology Group (RTOG) trials being performed at the time, primarily because of weight loss or comorbidity.

Thus, there are good reasons to be concerned that the efficacy established by randomized controlled trials may not translate into an identical level of effectiveness in the general population. Some empirical studies have called into question whether adherence to guidelines is associated with better outcomes. Allen et al. showed no difference in survival or mortality rates between patients who were and were not treated according to NCCN-defined surgical treatment guidelines. In a study of patients with stage I–III NSCLC, Duggan et al. noted that patients who were treated in accordance with prevailing Australian guidelines experienced slightly better outcomes than those who did not receive guideline-recommended treatment, but that trend was not significant.

It has therefore been suggested that, after evidence-based guidelines have been adopted, phase IV population-based outcome studies should be done to confirm the value of the new treatment in the general population. The results of a recent Canadian study were surprisingly reassuring. Following the publication of the positive results of a Canadian trial of adjuvant chemotherapy for resected NSCLC, this practice was rapidly adopted in Canada. Booth et al. evaluated the results of a phase IV population-based study that documented the rapid adoption of adjuvant chemotherapy in Ontario and reported that such therapy was associated with an increase in survival commensurate with that expected on the basis of the results of the preceding randomized controlled trial. The fact that the results of this particular trial were reproduced in routine practice does not guarantee the generalizability of the results of other randomized controlled trials, and additional phase IV population-based outcome studies are required to evaluate the societal benefit of the adoption of other promising new treatments.

To achieve optimal outcomes at the societal level, every patient must receive the correct treatment, and that treatment must be delivered correctly. Thus, in addition to practice guidelines for the selection of the appropriate treatment, additional standards are required to ensure the quality and accessibility of treatment.

General Quality Standards

In order to provide the optimal quality of treatment, the necessary structures and processes must be in place to deliver the treatment correctly. The term structure is used here to include human resources, physical resources, and organizational resources, and the term process includes all of the activities and procedures used to ensure the quality of treatment, as described earlier. Although practice guidelines target precisely defined subgroups of patients, the structures and processes required to ensure optimal quality of care are often applicable to much broader groups of patients. The same facilities usually serve the needs of diverse groups of patients, and the structures and processes that determine the quality of care are often common to many different types of cancer. Thus, the quality of the care that patients with lung cancer receive is in large measure determined by the degree of institutional adherence to general standards of practice that are applicable to the care of every patient with cancer.

The American College of Surgeons has been a leader in this field. In 1930, its Committee on the Treatment of Malignant Diseases released its first set of cancer program standards and created an accreditation program to evaluate the performance of a cancer clinic against those standards. As the management of cancer became increasingly multidisciplinary, the membership of this committee was broadened to include individuals from nonsurgical disciplines and its name was changed to the Commission on Cancer. Today, it provides a suite of program standards aimed at ensuring comprehensive, patient-centered, high-quality, multidisciplinary care for all patients with cancer.

Multidisciplinary Team Management

No individual specialty has all of the knowledge and expertise necessary for optimal decision-making related to the management of a complex disease in the modern era. The primary rationale for introducing MDTs is to bring together the expertise of all of the key professional groups in making clinical decisions for individual patients. Treatment decisions are made by consensus, reducing the risk that the bias of any individual physician will determine the final decision. There is usually considerable overlap in expertise among team members, providing a built-in opportunity for peer review of treatment decisions. The organized and open decision-making process provides a suitable forum for applying treatment guidelines and for identifying patients who are eligible for participation in clinical trials. When several different health professionals are involved in the overall plan of care for the individual patient, the team structure fosters communication among all those involved. MDT management therefore has been recommended as a mechanism for improving the quality of care for patients with various complex diseases, including diabetes, stroke, ischemic heart disease, and cancer.

On the basis of the persuasive arguments in favor of its use, the MDT approach has been widely adopted in cancer care systems throughout much of Europe, the United States, and Australia. At the outset, there was very little empirical evidence that this approach actually improved the quality of patient care or the outcomes of treatment. Although some observers remain skeptical about the value of MDT work, accumulating evidence suggests that this approach does lead to better decision-making and that it may improve outcomes, at least for patients with certain types of cancer. In one study from the United Kingdom, the vast majority of health professionals reported that they enjoyed working within the framework of an MDT and many reported that doing so had increased their job satisfaction.

Multidisciplinary Team Management of Lung Cancer

MDT management of lung cancer is particularly important as studies have shown that practice varies widely, that physician views are not always concordant with guidelines, and that the different specialists involved in treatment tend to be biased in favor of their own modality of treatment. It is also essential in the setting of lung cancer because multimodality treatment is common, pathologic subtyping is continually evolving, and patient comorbidities can have a substantial impact on the safety of therapeutic options.

The MDT should involve all of the clinicians who are involved in the diagnosis and treatment of lung cancer, including respiratory physicians, cardiothoracic surgeons, medical oncologists, radiation oncologists, palliative care physicians, and lung cancer nurses. The presence of a pathologist, radiologist, and nuclear medicine physician is also essential for accurate interpretation of the pathologic and imaging findings that underpin management recommendations. Ideally, all patients should be discussed at an MDM where all of these disciplines are represented. The presence of clinicians from different specialties should serve to reduce specialty bias in the treatment of patients and to inform colleagues of the role of different treatment modalities. The potential benefits of an MDM are increased compliance with evidence-based guidelines, improved utilization of treatment, increased referrals for psychosocial care, improved timeliness of treatment, and increased recruitment to clinical trials.

The impact of an MDM ideally would be tested in a randomized controlled trial, but this type of study is difficult to conduct for such a complex intervention. However, a number of investigators have sought to evaluate the impact of MDMs either by comparing patterns of care before and after the implementation of an MDM or by comparing cases discussed in an MDM with contemporaneous controls who were not discussed in the MDM. Both of these approaches are susceptible to bias. The longitudinal “before-and-after” design is vulnerable to changes in case mix, staging, and management over time, whereas the use of contemporaneous controls is susceptible to bias in the selection of patients for the MDM. In analyzing the results of such observational studies, it is therefore important to control, as much as possible, other factors that may influence the choice of treatment or its outcome.

Forrest et al. evaluated the impact of the introduction of a multidisciplinary lung cancer team on the treatment of patients with inoperable NSCLC. The pre-MDM data were collected retrospectively for 1997, the year preceding introduction of the MDM, and the post-MDM data were collected prospectively for 2001. The authors found a significant increase in formal staging of lung cancer (81% vs. 70%; p = 0.04), a significant increase in the use of chemotherapy (23% vs. 7%; p < 0.001), a decrease in the use of palliative care alone (58% vs. 44%; p = 0.05), and no significant change in the use of radiotherapy. The median survival was significantly greater for patients who were discussed at the MDM (6.6 vs. 3.2 months; p < 0.001). The two cohorts were not balanced for stage, with the post-MDM group including fewer patients with stage IIIA disease, which one would expect to result in poorer survival.

The improvement in survival therefore was attributed to the increased use of chemotherapy. However, it seems somewhat improbable that this factor alone could have been responsible for the large difference in survival. It remains possible that differences in case mix between the groups, not fully controlled for in the analysis, may have contributed to the difference in survival.

Erridge et al. compared patterns of care in Scotland before and after the implementation of a number of changes related to the treatment of lung cancer, including the introduction of MDMs at which all patients with newly diagnosed lung cancer were discussed, the introduction of management guidelines, and an increase in the number of oncologists specializing in the treatment of lung cancer. Although the overall active treatment rate was unchanged, the investigators found a significant increase in the use of curative radiotherapy (5% vs. 15%; p < 0.001) and chemotherapy (7% vs. 18%; p < 0.001) for patients with NSCLC. The median survival time improved from 4.1 to 5.2 months ( p = 0.004). As there were multiple concurrent changes in cancer care over time, it was not possible to tease out the impact of MDMs alone.

Seeber et al. compared the treatment of lung cancer before and after the implementation of an MDM videoconference that was held at a peripheral site. The investigators found that the MDM changed the treatment recommendations of the presenting team in 25% of the cases. The use of radiotherapy increased from 30% to 70% ( p = 0.001) following the introduction of the MDM. There was no change in the use of chemotherapy, and the authors did not comment on the use of surgery.

The largest study to evaluate the impact of the introduction of an MDM at a single institution was performed by Freeman et al., who compared the care of 535 patients who were treated before the implementation of an MDM with that of 687 patients who were treated after its introduction. The investigators found increased completeness of staging (93% vs. 79%; p < 0.0001), greater adherence to NCCN guidelines (97% vs. 81%, p < 0.0001), and reduced time from diagnosis to treatment (17 vs. 29 days; p < 0.0001) after the introduction of the MDM. There was also greater use of neoadjuvant chemotherapy and surgical resections for patients with stage IIIA NSCLC who were treated after the introduction of the MDM.

Two studies have compared the treatment of and outcomes for patients who were discussed at an MDM with those who were not in a cohort of patients who were seen at the same center over the same time period. Bydder et al., in a small study of patients with inoperable NSCLC, found no difference between the MDM and non-MDM subgroups in terms of treatment but reported that the MDM subgroup had a longer median survival time (280 vs. 205 days; p = 0.05). However, the authors noted some imbalances between the groups in terms of tumor characteristics and acknowledged that selection bias may have affected the results. In a much larger study of 988 patients who were diagnosed with lung cancer between 2005 and 2008, Boxer et al. compared 504 patients who were discussed at the MDM with 484 patients who were not. The investigators found that patients in the MDM subgroup were significantly more likely to be treated with radiotherapy (66% vs. 33%, p < 0.001), to be treated with chemotherapy (46% vs. 29%, p < 0.001), and to be referred for palliative care (66% vs. 53%, p < 0.001). The groups did not differ with respect to the rate of surgical treatment. Multivariate analysis showed that discussion at an MDM was an independent predictor of nonsurgical treatment and palliative care referrals but was not predictive of survival. Clinician selection bias may have had some influence on these results. Patients with better performance status may have been selected for presentation at the MDM on the assumption that they would receive treatment. Unfortunately, ECOG performance status was not available for the patients in the non-MDM cohort, and the authors therefore were unable to control for this factor in the analysis. The authors tried to minimize potential bias by assigning population-derived ECOG data to compensate for this limitation, but the potential for confounding remains.

Single-arm studies also may provide some useful information about the quality of care for patients who are discussed in an MDM. Conron et al. described the activity of a lung cancer multidisciplinary clinic at which 431 patients were discussed between 2002 and 2004. Management was compared with prospectively identified measures of quality of care. They found that 98% of patients with stage I to stage IIIA NSCLC have macroscopically complete surgical resection, that 100% of patients with stage IIIB NSCLC have positron emission tomography before curative chemoradiation therapy, that 84% of patients with stage IIIB NSCLC complete curative radiotherapy, that 86% of patients with stage IV NSCLC are referred for palliative chemotherapy, and that 100% and 85% of patients with limited-stage SCLC have complete staging and receive thoracic radiotherapy, respectively. In a study that was described earlier, Vinod et al. found 71% concordance between MDM recommendations and guidelines, but, in a follow-up study, Boxer et al. found that only 51% of patients actually received guideline-based care.

The efficacy of MDMs has been evaluated in terms of whether recommendations are translated into practice and whether deviations from these recommendations affect outcomes. Leo et al. analyzed the concordance between MDM-planned treatment and administered treatment in a study involving a cohort of French patients who were discussed in 2003 and 2004. Patients who did not receive the recommended treatment had poorer survival, although this finding did not reach significance. In the United States, Osarogiagbon et al. compared the outcomes for patients with any thoracic malignancy who received treatment that was concordant or discordant with MDM recommendations. Patients who received concordant treatment had a shorter time to clinical intervention (14 vs. 25 days, p < 0.002) and better median survival (2.1 vs. 1.3 years, p < 0.01). The authors were unable to identify the reasons for discordance in their study; however, Black race and lack of medical insurance were the two factors associated with the receipt of discordant care.

In summary, although there is no level 1 evidence to support the use of MDMs, observational studies have shown that patients who are discussed at an MDM are more likely to receive treatment (particularly radiotherapy and chemotherapy), more likely to receive potentially curative treatment, and more likely to be referred for palliative care. The increased use of all and any of these modalities, including palliative care, has the potential to improve survival, although there is little direct evidence that discussion at an MDM is associated with improved survival. For these reasons, case discussion at an MDM has been adopted as an indicator of quality of lung cancer care in some jurisdictions.

Although it is now widely accepted that multidisciplinary management is effective for improving the quality of care for patients with lung cancer, the optimal structure and operations of the MDT have yet to be defined. Taylor and Ramirez, in a study commissioned by the National Cancer Action Team in the United Kingdom, recently surveyed members of multidisciplinary cancer teams, with three objectives: (1) to identify the attributes of an effective MDT, (2) to learn how best to measure MDT effectiveness, and (3) to determine what support or tools MDTs need to be most effective. More than 2000 MDT members responded to the survey, of whom 53% were physicians, 26% were nurses, and 15% were MDT coordinators. The results of the survey demonstrated a high degree of consensus about the domains that are important for effective functioning of the MDT, including the membership, leadership, and governance of the team; the physical environment of the venue, technologic resources, preparation for meetings, administration of meetings, and attendance at meetings; the decision-making process, case management, and coordination of service; data collection, analysis, and auditing of outcomes; and development and training of participants. This very useful report provides many additional details of the elements of an effective MDT from the perspective of experienced team members. It should be essential reading for any member of a cancer program who is contemplating the introduction of an MDT.

If MDT care is worth doing, it is worth doing well. In England, the Improving Lung Cancer Outcomes Project brings together multidisciplinary health-care teams from different regions under the leadership of the Royal College of Physicians. A recent report from this group described a quality-improvement exercise in which each of 30 randomly selected MDTs were paired with one another in visiting the others’ services for a day, attending the MDM, and reviewing audit results. The most commonly identified problems concerned the way in which MDMs operated, including deficiencies in the amount of information available at the meetings, the way in which decisions were made, and methods for capturing outcomes. The teams then used standard quality-improvement methods to target specific problems that were identified in the peer-review process. Ultimately, the impact of this peer-review process will be evaluated by comparing the outcomes achieved by the 30 MDTs that participated in the intervention with outcomes achieved by MDTs that did not. This exercise may prove to be a very useful approach for enhancing the effectiveness of MDT care in the future.

MDT management necessarily generates added costs, and the cost-effectiveness of this approach has not been well evaluated. One small study from the United Kingdom, which considered only the salaries of senior staff and excluded preparation time and management costs, estimated the cost of discussing each patient to be £36.6 (US$60.64), whereas another study from the United Kingdom, which considered all costs, including use of audiovisual equipment, clerical time, preparation time for radiographic and pathologic examinations, and room overhead, estimated the cost per treatment plan to be £87.41 (US$144.83). If multidisciplinary management really results in improvements in treatment and outcome, then MDMs may still be very cost-effective. However, the cost of MDMs is clearly not trivial and further research on MDT management should seek to find ways to enhance its efficiency as well as its effectiveness.

Modality-Specific Quality Standards for Cancer Programs

Standards that target the quality of the major modalities of cancer treatment are also necessary to ensure the optimal treatment of lung cancer. For example, both Australia and Canada have established quality standards for radiotherapy programs, which are clearly pertinent to the optimal care of patients who have lung cancer and other malignant diseases. While these guidelines also deal with general aspects of patient care, they are most important for prescribing the structures and processes required for the safe delivery of radiotherapy, which are not stipulated in more general cancer program guidelines.

Standards for Acceptable Patient Volumes for Cancer Programs

The results of recent studies suggest that the outcomes of cancer may be better among patients who are treated at larger institutions. The initial observations were made in the field of cancer surgery, in which many studies have shown an inverse association between surgical mortality and the number of operations performed in a given hospital or by a given surgeon. von Meyenfeldt et al. recently performed a meta-analysis to explore the relationship between volume and outcome in the surgical treatment of lung cancer. The 19 studies that were included in the analysis proved to be very heterogeneous, particularly with respect to the definition of volume categories. However, pooled estimates showed a significantly lower surgical mortality in favor of larger hospitals (odds ratio, 0.71; 95% confidence interval [CI], 0.62–0.81) but no significant difference in terms of long-term survival (odds ratio, 0.93; 95% CI, 0.84–1.03). In a subsequent study of 4460 patients who underwent surgery for the treatment of lung cancer at 436 hospitals in the United States, Kozower and Stukenborg compared three alternative measures of volume for evaluating the volume-outcome relationship. The authors found no significant association between hospital procedure volume and in-hospital mortality when volume was measured as a continuous variable. A significant relationship was found when volumes were categorized into quintiles, but the magnitude of the association was small. The authors concluded that the apparent impact of hospital volume on mortality is dependent on how volume is defined, that volume is not consistently related to mortality, and that volume should not be used as a proxy measure for the quality of surgery. However, Lüchtenborg et al. recently found a strong and significant association between institutional volume and survival in a study from England involving 134,293 patients who had been diagnosed with NSCLC between 2004 and 2008, of whom 12,862 (9.6%) underwent surgical resection. The authors found that rates of resection are higher at high-volume hospitals, where surgery is performed more often on patients who were older and had more comorbidities. Despite these findings, survival was significantly better at hospitals at which more than 150 surgical resections were performed each year than at those at which fewer than 70 surgical resections were performed each year (hazard ratio, 0.78; 95% CI, 0.67–0.90).

There have been far fewer studies investigating the impact of case volume on the outcomes of radiotherapy, but there are good reasons to expect that a similar relationship may be found, particularly in the context of complex types of treatment. Lee et al. combined the results of two RTOG trials (RTOG 91-06 and RTOG 92-04) to address this question in the context of chemoradiation therapy for inoperable lung cancer. After controlling for other prognostic factors as much as possible, the investigators found significantly better survival among patients who were treated at institutions that enrolled five or more patients each year as compared with those who were treated at lower-volume centers (31% vs. 13% at 3 years). The magnitude of this volume effect did not diminish over the study period, and the authors concluded that collective institutional experience contributed more to the difference in outcome than the learning curve effect did. The investigators did not attempt to distinguish the effect of the experience of the individual oncologist from institutional experience.

These so-called volume effects are potentially very important. If they are real, they represent an extraordinary opportunity for improving the outcomes for patients with lung cancer. Intervention studies are now urgently required to confirm that limiting treatment to high-volume facilities improves overall outcomes. Further explanatory studies are also required to try to identify the underlying differences in care that are responsible for the differences in outcomes between low- and high-volume centers. If these causative factors can be identified, then it may be possible to develop strategies to increase the quality of care at smaller centers to the level provided at the larger centers, thus avoiding the need for centralization of services, which may come at the cost of decreased accessibility.