Infrastructure of radiation oncology in France: A large survey of evolution of external beam radiotherapy practice

Int. J. Radiation Oncology Biol. Phys., Vol. 61, No. 2, pp. 507–516, 2005 Copyright © 2005 Elsevier Inc. Printed in the USA. All rights reserved 0360-3016/05/$–see front matter

doi:10.1016/j.ijrobp.2004.06.009

CLINICAL INVESTIGATION

Radiation Oncology Practice

INFRASTRUCTURE OF RADIATION ONCOLOGY IN FRANCE: A LARGE SURVEY OF EVOLUTION OF EXTERNAL BEAM RADIOTHERAPY PRACTICE SOPHIE RUGGIERI-PIGNON, M.D.,* THIERRY PIGNON, M.D., PH.D.,† MICHEL MARTY, M.D.,‡ MARIE-HÉLÈNE RODDE-DUNET, M.D.,‡ BRIGITTE DESTEMBERT, M.D.,§ AND BÉATRICE FRITSCH, M.D.¶ *Caisse Nationale d’Assurance Maladie des Travailleurs Salariés, Direction Regionale du Sud Est, Marseille, France; †Department of Oncology Radiotherapy, Hôpital de la Timone, Université de la Méditerrannée, Marseille, France; ‡Caisse Nationale d’Assurance Maladie des Travailleurs Salariés, Direction du Service Médical, Paris, France; §Caisse Nationale d’Assurance Maladie des Travailleurs Salariés, Echelon local du Service Médical, Orléans, France; ¶Caisse Nationale d’Assurance Maladie des Travailleurs Salariés, Echelon local du Service Médical, Strasbourg, France Purpose: To study the structural characteristics of radiation oncology facilities for France and to examine how technological evolutions had to be taken into account in terms of accessibility and costs. This study was initiated by the three health care financing administrations that cover health care costs for the French population. The needs of the population in terms of the geographic distribution of the facilities were also investigated. The endpoint was to make proposals to enable an evolution of the practice of radiotherapy (RT) in France. Methods and Materials: A survey designed by a multidisciplinary committee was distributed in all RT facilities to collect data on treatment machines, other equipment, personnel, new patients, and new treatments. Medical advisors ensured site visits in each facility. The data were validated at the regional level and aggregated at the national level for analysis. Results: A total of 357 machines had been installed in 179 facilities: 270 linear accelerators and 87 cobalt units. The distribution of facilities and megavoltage units per million inhabitants over the country was good, although some disparities existed between areas. It appeared that most megavoltage units had not benefited from technological innovation, because 25% of the cobalt units and 57% of the linear accelerators were between 6 and 15 years old. Computed tomography access for treatment preparation was not sufficient, and complete data management systems were scarce (15% of facilities). Seven centers had no treatment planning system. Electronic portal imaging devices were available in 44.7% of RT centers and in vivo dosimetry in 35%. A lack of physicians and medical physicists was observed; consequently, the workload exceeded the normal standard recommended by the French White Book. Discrepancies were found between the number of patients treated per machine per year in each area (range, 244.5– 604). Most treatments were delivered in smaller facilities (61.6%). Conclusion. On the basis of the findings of this study, measures were taken to update the infrastructure of RT in France. A first evaluation showed an improvement of care supply in RT in the country. © 2005 Elsevier Inc. Radiotherapy, Facility survey, Infrastructure, Personnel, Activity.

INTRODUCTION

ment systems, computed tomography [CT] scanners) in terms of accessibility and costs. Between 145,000 and 160,000 new patients with cancer are treated each year by RT in France. In addition, because of the aging population, the incidence of cancer is rising steadily, particularly those treated by RT such as breast, prostate, and lung cancer (1). In France, authorization for the purchase of RT apparatus is subject to a ministerial decree dating from 1986 that limits the number of machines to 6 of 106 inhabitants (1/167,000 inhabitants, Sanitary Map Index). Because of the health organization in the country, practicing RT can be different in relation to payment and financing. In public

In 1999, the French government announced a cancer plan for the 2000 –2005 period. It was designed to define the new policy for Regional Organization of Health in oncology, in particular to organize the care supply among the different country health structures at the regional level. The aim was to provide an equal chance for patients to have access to an optimal cancer treatment in each country area. It was then necessary to rethink the place of radiotherapy (RT) in this new organization and how to take into account technological evolution (e.g., multileaf collimators, electronic portal imaging, electronic administrative and image-data manageReprint requests to: Sophie Ruggieri-Pignon, M.D., Caisse Nationale d’Assurance Maladie des Travailleurs Salariés, Direction Regionale du Sud Est, 195 Boulevard Chave, Marseille 13005, France. Tel: (⫹33) 49-185-8688; Fax: (⫹33) 49-142-4401; E-

mail: [email protected] Received Feb 9, 2004, and in revised form Jun 3, 2004. Accepted for publication Jun 4, 2004. 507

508

I. J. Radiation Oncology

● Biology ● Physics

hospitals, financing comes from the global hospital endowment (DG hospitals) and practitioners are not paid on a per-procedure basis. In private for-profit hospitals (NGAP hospitals), the payment of RT activity relies on a scale value resulting from the combination of a base value designated by the Z letter and multiplication factors supposed to measure the medical cost of each RT procedure. The value of one “Z” is 1.67 euros. During the study period, the technical preparation and quality assurance procedures (dosimetry, portal imaging, in vivo dosimetry) accounted for only a small part of the price of a treatment and the amount depended essentially on the given dose and the type of machine. The remuneration basis for NGAP hospitals, dating from 1974, did not take into account the new technology. The investment and replacement of material was ensured by private funds. Many issues can be raised concerning the optimal planning of facilities, equipment, and manpower for treating patients with the best possible efficiency. First, was the number of megavoltage units, treatment capacity, and geographic distribution of facilities optimal in the country? Second, how could one predict the impact of rapid technological evolution on the functional organization of RT and at what intervals should the present installations be replaced? In addition to an appreciation of population needs, many factors had to be taken into account to answer these questions (2, 3). A study of technical performance and the technical environment of existing megavoltage units was required to assess the necessity for replacing treatment installations. Professionals pointed to an increase in the activity of centers that resulted in not allowing all patients equal access to modern techniques (4). In addition, a rise in additional megavoltage unit authorizations would be mandatory to absorb the large patient load per machine, which is incompatible with these more time-consuming technologies. However, the potentially low rate of radiation therapists entering the profession and the lack of medical physicists would moderate this demand, because their role in these new technological opportunities and quality assurance programs is likely to rise. In addition, if growing costs were justified individually with regard to local control amelioration and side-effect limitation, from a collective viewpoint, it was necessary to verify whether the expected benefit for the population from these techniques would be consistent with the increase in costs. Therefore, financing possibilities, taking into account budget consequences, are a decisive factor for technical evolution of RT in France, as it is in other countries (4). For NGAP facilities, investment capacities had to be evaluated with respect to their modality of payment. For a DG hospital, only a political decision could change the situation. Consequently, the three health care financing administrations that cover health care costs for the French population decided to undertake a comprehensive survey to evaluate the needs of the population and to make proposals to enable an evolution of the practice of external beam RT in France.

Volume 61, Number 2, 2005

The aim of the present report was to describe the findings of that survey and describe the measures taken at its outcome. METHODS AND MATERIALS This study was a part of the National Intergroup Program of the three health care financing administrations (Caisse Nationale d’Assurance Maladie des Travailleurs Salariés, Mutuelle Sociale Agricole, and Caisse Nationale d’Assurance Maladie des Professions Indépendantes) that cover health care costs for nearly 100% of the French population. It was designed and conducted by a work group of medical advisors.

Quality control procedure The data were entered on a standard collection form designed by a multidisciplinary committee of radio-oncology specialists, one medical oncologist, one surgeon, medical physicists, and public health specialists. The aim of this expert committee was also to conduct training for physicians who had the responsibility for the data collection. In addition, they reviewed, evaluated, and validated the findings and results.

Data collection A comprehensive list of public and private centers performing RT was compiled from the Health Ministry. The type of hospital was recorded (university hospital, non– university-affiliated public hospital, or private for-profit institution). Site visits were conducted to each facility by medical advisors between October and December 1999. They collected data after being trained in RT. Part 1 of the questionnaire included questions concerning the activity and equipment of the RT center (apparatus, dosimetry system, and devices). Part 2 of the questionnaire was dedicated to means related to the structure of treatment. Part 3 focused on treatment strategy and technical procedure of a patient sample randomly chosen in each center, and part 4 covered the overall costs (i.e., machine, dosimetry, buildings, personnel, consumables, furniture, management). The purpose of the site visit was to obtain data for the general treatment planning survey and to review patient charts to extract data for the patient survey. The general treatmentplanning questionnaire was given to the facility’s radiation oncologist at the beginning of the site visit with the request that it be completed. Variables concerning equipment and personnel were taken into account as of January 1, 2000. Data measured over a period, such as activity, were requested for the calendar year 1998 or the first 9 months of 1999. Criteria for activity evaluation were the number of new treatments, the number of new patients, and the number of sessions. The results of the disease site treatment planning surveys will be reported elsewhere; this report describes the results of the general planning survey addressing non– diseasespecific elements, such as equipment, personnel, and activities. These results were compared to the recommendations produced by professionals for RT practice and quality assurance and published in the French White Book (5).

Statistical analysis A separate analysis was performed for DG financing institutions (university hospital and non– university-affiliated public hospital) and NGAP financing hospitals (private for-profit institutions). The variables were cross-checked at the regional level with particular attention to the completeness of responses. The number of facilities and megavoltage units were compared with the ministry data.

Infrastructure of RT in France

● S. RUGGIERI-PIGNON et al.

509

Table 1. Distribution of installations according to region and type of facility

Region

Facilities (n)

Facilities/million inhabitants (n)

Linear accelerators (n)

Alsace Aquitaine Auvergne Basse Normandie Bourgogne Bretagne Centre Champagne-Ardennes Corse Franche Comté Haute Normandie Ile de France Languedoc-Roussillon Lorraine Limousin Midi-Pyrénées Nord Outre Mer Provence Pays de Loire Picardie Poitou Charente Rhône Alpes Total

4 10 6 4 6 8 8 5 2 3 4 28 10 6 4 7 11 4 12 8 6 6 17 179

2.31 3.44 4.59 2.81 3.73 2.75 3.28 3.73 7.69 2.69 2.25 2.56 4.36 2.60 5.63 2.74 2.75 2.40 2.66 2.48 3.23 3.66 3.01 Mean 3.36

5 15 6 7 7 13 10 8 1 5 6 50 14 12 4 11 17 4 20 16 6 6 27 270

Coherence between the type of institution and their status noted on the CNAM listing was verified. Extreme values for activities were pointed out and were the subject of additional investigation. Missing or aberrant data were collected by a regional project director who sent queries to medical advisors for on-site verification. After definitive validation with Microsoft Excel and SPSS software, aggregation of all data were made at the national level where statistical analysis was performed with SPSS software.

RESULTS The study concerned 179 RT facilities of the 185 indexed by the ministry. Two centers were excluded because of their specificity (the military hospital in Paris and the proton therapy center in Orsay). In addition, two others were definitively closed and two were authorized but not yet installed. NGAP financing centers accounted for 53.6% and DG financing for 46.4% but more than one-half of the machines were in this second group. Distribution of centers In these 179 institutions, 357 machines were installed: 270 linear accelerators and 87 telecobalt units. Table 1 shows the distribution of these installations according to region and type of facility. The regional average of facility number per million inhabitants was 3.36. Three regions were significantly over the mean (Auvergne, 4.59; Limousin, 5.63; and Corse, 7.69) and two were under (Alsace, 2.31; and Haute Normandie, 2.25). Despite the Sanitary

Cobalt units (n) 4 3 3 1 2 4 3 3 1 1 4 22 1 2 3 4 5 3 6 2 3 4 3 87

Total (n)

Megavoltage units/million inhabitants (n)

Inhabitants 1999 (n)

9 18 9 8 9 17 13 11 2 6 10 72 15 14 7 15 22 7 26 18 9 10 30 357

5.19 6.19 6.88 5.63 5.59 5.85 5.33 8.20 7.69 5.37 5.62 6.57 6.54 6.06 9.86 5.88 5.51 4.2 5.77 5.59 4.85 6.10 5.31 Mean 6.08

1,734,145 2,908,359 1,308,878 1,422,193 1,610,067 2,906,197 2,440,329 1,342,363 260,196 1,117,059 1,780,192 10,952,011 2,295,648 2,310,376 710,939 2,551,687 3,996,588 1,667,436 4,506,151 3,222,061 1,857,834 1,640,068 5,645,407 60,186,184

Map Index, disparities were found when considering the number of megavoltage units per million inhabitants. If Limousin (9.86), Champagne-Ardennes (8.20), and Corse (7.69) were over the index, Picardie (4.85) and the overseas departments (4.20) were underequipped. However, no “desert area” was present in the country, because the percentage of patients treated in centers further than 100 km from their residence was only 5.5%. After exclusion of patients treated in distant centers when a closer center existed, the percentage fell to 1.2%. Number of machines and energy A total of 65 centers had one megavoltage unit exclusively; 78 facilities had two machines and 25 had three. The 14 centers with four or more megavoltage units accounted for 64 machines. Cobalt units represented 25% of the total and were the oldest apparatus; 48% were ⬎15 years old and 43% were between 6 and 15 years old. Linear accelerators accounted for 75%. Only 7% were older than 15 years and 57% were between 6 and 15 years of age. A percentage of 36% of all linear accelerators were ⬍6 years old. Because only accelerators installed after 1994 could benefit from the technological innovations, most machines were obsolete. In addition, 29 of 77 centers had a cobalt source too weak for optimal use (time of irradiation ⬎1.50 min calculated for 1 Gy, field size 10 cm ⫻ 10 cm, depth 5 cm, and source to skin distance [SSD] 80 cm). Cobalt units were less frequent in facilities with three

510

I. J. Radiation Oncology

● Biology ● Physics

Volume 61, Number 2, 2005

Table 2. Photon energies and level of equipment of facilities in relation to number of megavoltage units Megavoltage units/center (n) Variable

1

2

3

ⱖ4

Centers (n) Photon energy (n) ⬍10 MV 1 ⱕ6 MV and 1 ⱖ15 MV 1 of 8–10 MV plus 1 ⱖ15 MV Simulators (n) TPS NB centers 2D 3D Network Image data Technical data Customized immobilization device (n) Multileaf collimator (n) Tailored block (n) Quality control (n) X-ray films EPID

62

78

25

14

24 (38.7) 37 (59.7) 1 (1.6) 53 (85.5)

8 (10.3) 64 (82) 6 (7.7) 77 (98.7)

0 25 0 25 (100)

0 14 0 14 (100)

57 (92) 31 34 (55)

76 (97.4) 76 60 (77)

25 (100)

14 (100)

25

14

24 (39) 11 (17.7) 52 (83.9)

48 (61.5) 31 (39.7) 71 (91)

19 (76) 13 (52) 25 (100)

11 (78.5) 9 (64) 14 (100)

4 55

16 76

12 25

9 14

61 14

78 37

25 18

14 11

Abbreviations: TPS ⫽ treatment planning system; 2D ⫽ two-dimensional; 3D ⫽ three-dimensional; EPID ⫽ electronic portal imaging device. Data in parentheses are percentages.

machines (13.3%) and accounted for 30% of units in centers with four or more megavoltage units. DG financing facilities were equally distributed in relation to the number of RT units. NGAP centers were mainly small, with 90% having one or two machines. Twenty-one centers (11.7%) had photon energy ⱕ6 MV (accelerator and cobalt). Energies ⬎18 MV were present in 53.6% of centers, predominantly in NGAP financing centers (60.4%). Table 2 shows the distribution of photon energies in relation to the number of megavoltage units per facility.

Nearly 18% of facilities, equally distributed in NGAP and DG financing, had only low energy (⬍10 MV). Of these 32 centers, 8 had two machines. Only 78% had a minimum of two photon energies (one ⱕ6 MV and one ⱖ15 MV), as recommended by professionals. Only one facility with an accelerator did not lay out electrons. Most facilities had a large choice of electron energies, with clinical coherence, except for two (one with one treatment unit and one with two), which had only electron ⱕ10-MeV energies (Table 3).

Table 3. Electron energies in relation to number of megavoltage units per facility and type of financing hospital Megavoltage units (n) Variable Facility type Available energies (n) ⱕ3 4–7 ⬎7 Available energies (MeV) ⱕ10 10–15 ⬎15 ⱕ10 and 10–15 ⱕ10 MEV

1

2

DG (n ⫽ 26)

NGAP (n ⫽ 36)

DG (n ⫽ 28)

NGAP (n ⫽ 50)

DG (n ⫽ 15)

NGAP (n ⫽ 10)

DG (n ⫽ 14)

0 14 3

2 17 9

0 11 17

2 28 20

0 4 11

0 3 7

1 13

12

22

26

44

14

8

14

5 0

5 1

2 0

4 1

1 0

2 0

0 0

Abbreviations: DG ⫽ public hospital; NGAP ⫽ private for-profit hospital.

Total (n)

ⱖ4

3

NGAP (n ⫽ 0)

179

0

4 78 80

0

140 0 0

19 2

Infrastructure of RT in France

● S. RUGGIERI-PIGNON et al.

511

Table 4. French White Book recommendations for minimal equipment for radiotherapy facility and number of facilities in agreement with these recommendations Level of technicability 1

2

3

Technique type Indication

Simple Metastasis Locally advanced tumor

Sophisticated Exclusive irradiation Adjuvant irradiation

High technicity Total body irradiation Total skin irradiation Radiosurgery Pediatric irradiation IMRT

Anatomic data Simulation Dosimetry Dose–volume histogram In vivo dosimetry Data management system Facilities (n) 1 unit 2 units 3 units ⱖ4 units

Simple material Yes 2D No

CT or MRI Yes 3D Yes Yes Yes

Yes Yes

0 7 7 7

0 1 5 6

Abbreviation: IMRT ⫽ intensity-modulated radiotherapy.

Technical facilities A CT facility for dosimetry was available in all centers except for four, which did not use CT scanning for dosimetry. However, the available time for use varied among the centers. Ten had a CT scanner dedicated to RT installed in the RT department. The vast majority of CT scans were obtained from CT scanners located outside the RT department but in the hospital. Overall, the mean time of access was 6 h/wk (range, 1– 40 h) with a median of 4 h. It depended on the size of the center: 3 h 30 min/wk for centers with one unit, 5 h for centers with two machines, 8 h for those with three units, and 16 h for centers with four or more units. An estimation of the CT scan time necessary for a RT center was proposed by experts according to two hypotheses. The minimal hypothesis accounted for 3 patients/h, and the maximal hypothesis accounted for 2 patients/h. It was assumed in this value that all patients underwent CT, but that the real need would be between these two estimates. The first estimate resulted in an average deficit of 3 h/wk (range, 2– 6 h) for 71 centers; the second estimate resulted in an average deficit of 6 h/wk (range, 3–11 h) for 110 centers. The deficit increased with the size of the center. Only 10 centers had no simulator; 58% stated that they performed virtual simulation. However, this last value was probably not true at that time. The presence of a threedimensional treatment planning system (TPS) does not guarantee its use; however, 75% of centers had that possibility. Seven centers had no TPS. For simplification, two types of network were identified: an image-data management system and a technical-data management system. The first was present in 102 centers (57%), 64 (36%) were equipped with the second, and both were present in 15%. This modern equipment was present in

facilities with the greater number of treatment units. Customized immobilization devices were used in 90% of centers for the head and neck, and 25% used expanding foam type immobilization for other tumor locations. Only nine centers did not use tailored blocks. Multileaf collimators were present in 41 centers for a minimum of one accelerator. An electronic portal imaging device was available in 80 centers (44%), but not for all treatment units in the same facility. In vivo dosimetry with ionization chambers, diodes, or thermoluminescent dosimetry was available in 63 institutions (35%), of which 56 were treating ⬎500 patients annually. In 1996, the French White Book recommended the minimal equipment for a RT facility, consisting of adapted immobilization devices; simulator and CT scan, CT scan simulator, or CT scan with virtual simulation; two-dimensional or three-dimensional TPS; and X-rays films for controls. Only 88 institutions (49.2%) met all these criteria. More recently, the recommendations were updated with three levels of technicality. With the latter recommendations, few institutions were deemed to have acceptable quality (Table 4). Personnel Table 5 displays the number of full-time or part-time radiotherapists and full-time equivalent (FTE) medical physicists and technologists. The number of radiotherapists grew with the number of treatment units, but some facilities with two treatment units had only 1 physician and one DG center with three treatment units worked with 2 radiotherapists. The number of FTE medical physicists did not follow the same progression as the number of physicians. Two facilities with only cobalt units had no medical physicist and

512

I. J. Radiation Oncology

● Biology ● Physics

Volume 61, Number 2, 2005

Table 5. Number of radiotherapists, medical physicists, and technologists in relation to number of megavoltage units per facility Megavoltage units/facility (n)

Physicians (n) Global NGAP DG Physicists (n) Global NGAP DG Technologists (n) Global NGAP DG

1

2

3

ⱖ4

2 (1–4) 2 (1–4) 2 (1–3)

3 (1–8) 3 (1–8) 3 (1–6)

5 (2–13) 6 (4–12) 5 (2–13)

8 (6–20) — 8 (6–20)

1 (0–2) 1 (1–2) 1 (0–2)

1 (1–3) 1 (1–3) 2 (1–3)

2 (1–4) 1 (1–3) 3 (1–4)

2.1 (2.8–6.5) — 2.1 (2.8–6.5)

4 (1–9) 4 (1–8) 4 (1–9)

7 (2–16) 6 (2–14) 8 (5–16)

14 (4–18) 11 (4–18) 14 (6–18)

17 (12–37) — 17 (12–37)

Data presented as median, with range in parentheses.

42 centers with two treatment units and 7 with three had only 1 FTE medical physicist. Eight facilities with four treatment units or more had 1.8 and 2.2 FTE medical physicists. A growing gap was found concerning FTE medical physicists between NGAP and DG institutions, in favor of the latter. On average, more medical physicists worked in facilities with the greater number of treatment units. The FTE number of technologists was greater in centers with more treatment units, but we noted a high degree of heterogeneity among centers with the same number of treatment units. The workload for radiotherapists, medical physicists, and technologists exceeded the normal standard as recommended by the White Book. Table 6 displays the number of facilities outside the recommendations. Activity Overall, 162,200 patients were treated in 1998. Nearly 40% were treated in facilities with two treatment units and 23% in centers with three treatment units. However, logically, the institutions with the greater number of megavolt-

age units received more patients. The average number of patients treated per machine was similar regardless of whether the center had one or three treatment units. The proportions were also close for centers with two or four treatment units. Although they accounted for fewer megavoltage units, NGAP institutions treated more than one-half of the patients (53.4%); these centers treated more patients per megavoltage unit. Also, a disparity was noted between the number of patients treated by machine in each area. The extreme values were 604 patients per machine in Provence and 245 patients per machine in Corse (Fig. 1). The greater number of sessions were found in facilities with two treatment units (42%), followed by centers with three (22%). This distribution corresponded to the patient numbers. The average number of 20 sessions/patient was similar for each type of institution according to the number of treatment units, but not according to the type of financing. NGAP facilities performed more sessions per machine than DG ones. The number of treatments was calculated from the data

Table 6. Number of centers out of the White Book recommendations concerning the number of radiotherapists, medical physicists, and radiation technologists

Radiation oncologists (RO) White Book Survey ⬍150 pts/y/RO ⬎350 pts/y/RO Medical physicists (MP) White Book Survey ⬍400 pts/y/MP ⬎600 pts/y/MP Technologists White Book Centers with ⬍2/unit

One megavoltage unit

Two megavoltage units

Three megavoltage units and more

1/200–350 pts/y

1/150–350 pts/y

1/150–250 pts/y

3 17

4 36

0 11

1 part-time minimum

1/400–600 pts/y

0 1

2 36

5 23

2 3

4 5

2/unit 6

Infrastructure of RT in France

from the first 9 months of 1999. Most of them were performed in smaller facilities (61.6%); centers with three or more units accounted for 38.4% of treatments. The mean waiting time for treatment, calculated from the data from 165 centers, was 15.3 days between the first consultation by a radiation therapist and the beginning of treatment. Except for facilities with only one treatment unit, the waiting time was always longer in a DG hospital than in private for-profit institutions: 19.2 vs. 14.3 days for two treatment units and 20.2 vs. 11.7 days for three. The 14 largest facilities, all in DG hospitals, had a mean waiting time of 14.5 days.

Treatment type Three classes were considered: primary tumor, metastasis, and benign disease. All types of facilities in relation to the number of treatment units treated all types of tumors. No specialization was found; in particular, even small centers with low-photon energies also treated deep tumor locations. Different types of facilities treated metastases in the same proportion. Benign disease accounted for mean of 3.7%;

● S. RUGGIERI-PIGNON et al.

513

however, 31 institutions treated a percentage ⬎8%. Fortysix facilities (8 NGAP and 38 DG) were using complex RT techniques: total body irradiation (n ⫽ 33), craniospinal irradiation (n ⫽ 26), cranial irradiation in stereotaxic condition (n ⫽ 21), total skin electron irradiation (n ⫽ 9), intraoperative RT (n ⫽ 14), and RT under general anesthesia (n ⫽ 12). Pediatric radiation oncology was performed in 50 facilities: 9 with one megavoltage unit, 18 with two, 10 with three, and 13 with four.

Quality assurance for megavoltage units Quality assurance has received much attention in the past few years for machines, TPS, simulators, film processors, and blocking systems. Our survey only asked questions about the capacity to give the number of calibrations per month and per machine, the number of beam quality control by year and by machine, and the control board for each treatment unit. Nine facilities were unable to answer these questions. Eighteen institutions did not respond positively to the three questions; of those, 14 failed to provide the number of calibrations.

Fig. 1. Mean number of patients per machine annually in each region.

514

I. J. Radiation Oncology

● Biology ● Physics

DISCUSSION No universal standard has been set for equipment rate, technicality level, use of facilities, and personnel for an optimal quality of RT. Each country determines its proper policy in relation to financing means, proper health organization, and priorities, which sometimes lead to severe arbitration between different health actors of different medical specialties. Additionally, recommendations, which can differ from one country to another (5– 8) are edited by professionals, but policymakers rarely follow them. In France, cancer has been recognized on several occasions as a public health problem. A cancer plan designed for the 2000 –2005 period was started again in March 2003 (9). It took into account the measures proposed for RT after this large survey was completed in 1999. This was the first study of the treatment planning structure undertaken in France. The aim was to examine the actual situation of RT in the country and to see which reasonable measures could be taken to enable an improvement in quality care in this field. These measures would encompass the interest of patients and professionals and meant that the government was ready to agree to reach this goal. In addition, it will serve as baseline for future studies. In radiation oncology, the technical quality of treatment does not summarize the quality of the therapeutic process. Accessibility to facilities, which also depends on their proximity and the frequency and length of displacement plus waiting time in centers, constitutes a major and immediate discomfort for patients (10). This problem is encountered when a centralized organization of facilities is preferred (11). In contrast, the decentralization option also has its disadvantages, including the risk of multiplication of facilities with only one treatment unit and relatively low activity, maintenance of treatment centers with limited technological levels because of the large investments, nomadism of some physicians and/or physicists with activity in several institutions, and multiplication of needs with scarce personnel, such as physicians and physicists. All these factors were observed at the end of this investigation. The distribution of facilities and megavoltage units per million inhabitants throughout the country was good, although some disparities existed between areas. However, these installations were mainly of small size, because 78% had one or two megavoltage units. This figure did not differ from that in other developed countries in which the number of machines per center was similar (12–15). Eleven percent of facilities had a load of ⬍250 patients/machine annually. This low activity contributed to a form of medical staff waste in regard to the weak French medical demography. It can be justified only for some structures situated in an isolated area. However, nearly one-third of facilities had excessive activity greater than the threshold recommended by the French White Book (300 –500 patients/machine) (5) or by the European Organization for Research and Treatment of Cancer (400 patients/machine) (6). A total of 32 centers treated ⬎600 patients/unit annually. It was particularly true for NGAP

Volume 61, Number 2, 2005

centers. Two reasons accounted for this situation. First, in areas with higher activity per megavoltage unit, the saturation of the Sanitary Map Index prohibited installation of additional machines. Second, the investment capability for NGAP centers was submitted to its profitability. This point perhaps had an influence on the quality of the machines and their technical environment, because a part of the megavoltage units had to be replaced. This was the case for the cobalt units, which represented 25% of the machines. One-half of them were ⬎15 years old. Sixteen were isolated installations. Seven percent of accelerators were also ⬎15 years old, and 57% were between 6 and 15 years old. Thus, most of the working megavoltage units did not benefit from technical innovations. It was also noted that CT scan access was not sufficient and that complete data management systems were not well developed (15% of facilities). If customized blocks were used in nearly all centers, the number of multileaf collimators was low. In addition, treatment quality control was suboptimal, because an electronic portal imaging device was present in only 44.7% of centers and in vivo dosimetry available in 35%. Although the level of equipment was very different among facilities, a relative homogeneity of tumor location treated was observed. Thus, a major investment effort appeared necessary to put all facilities in a good state for the realization of treatments of technicality level 1 and 2 as defined by French professionals. This led to a question on the relationship between the payment modalities for therapeutic acts and their consequences in terms of equipment and professional behavior (4). It was particularly true for NGAP facilities but also for public financing institutions. Professionals are induced to use equipment whether or not it is obsolete, but an inadequate economic incentive rarely permitted them to upgrade it. Another worrying issue was the lack of a sufficient number of oncologic radiotherapists and medical physicists in most facilities. When manpower levels were compared with the recommendations of the French White Book, 35% of facilities had overactivity for radiotherapists and 33% for medical physicists. A recently published report (16) concerning French medical demography emphasized the present and future deficit of physicians. Alleviating this problem depends on the enlargement of the entrance score at the medical university, on the capability of medical university to attract students toward radiation oncology, and on the improvement of radiotherapists’ payment. Although urgent, suggested measures did not give hope for an increase in new specialists before 10 –15 years. Measures have also been studied to improve the physicists’ status, particularly in public hospitals, to make the profession more attractive. The creation of the dosimetrist corporation, which does not exist in France, is envisaged. At the conclusion of this survey, the National Intergroup Program defined its position toward RT. First, the installation of additional treatment units has been justified in regard to the overactivity of numerous facilities and the foreseeable increase in patients in the future. The

Infrastructure of RT in France

priority is to reinforce centers developing the highest activity as long as their technological environment, medical oncology services, and available personnel satisfy all criteria required by the ministry recommendation for geographic oncology organization. At the same time, the replacement of old treatment units and the elimination of cobalt units must be undertaken. Second, an optimal distribution of facilities throughout the country in relation to demographic constraints, costs, and existing analysis has to be developed (2). The translation of facilities with only one treatment unit into facilities with two machines was the first goal—to be achieved, eventually, by regrouping small centers with low activity. Only facilities with at least two megavoltage units could be created, preferentially close to a hospital with confirmed oncology activity. The maintenance of isolated centers with only one megavoltage unit could be accepted only for a specific geographic consideration, on the sole condition that it works in connection with a reference center. However, an amelioration of quality policy and development of standardized procedures must necessarily accompany human and material investment. It was obvious that to realize these aims, two joined conditions are necessary: a deep change in the Sanitary Map Index, which was at the basis of megavoltage unit installation, and a modification of the RT Remuneration Acts (4). Thus, in November 2001, the modality of attribution of authorizations was changed. The new Sanitary Map Index was increased in each sanitary region (one treatment unit per 140 –165,000 inhabitants) in relation to the exigencies of modern RT, such as much more time for preparation and treatment planning. It was also modulated on a regional basis, depending on the geographic, demographic, and medical conditions. Finally, the Sanitary Map Index was definitively suppressed by September 2003 and the care supply in RT is now organized from epidemiologic data and selected criteria defined in each region with close participation of professional experts. The benefits of these measures will not be immediate, considering that after

● S. RUGGIERI-PIGNON et al.

515

agreement for a new machine, a delay of 3 years occurs before its installation. In the second step, the Payment of RT Acts was reformed. For a DG financing hospital, reinforcement of means for RT was encompassed in the 70 measures announced in the Cancer plan by March 2003. In private facilities, the Committee for the Nomenclature of Radiotherapy Acts modified, in June 2002, the payment of RT. Modern RT practices and new technologies were taken into account. The price was weighted in favor of preparation and dosimetry phases, with the detriment to the given dose. In addition, treatments with cobalt units were paid less. An overall sum was allocated for 2 years. At the end of this period, an evaluation will be performed to study the impact of these decisions on RT infrastructures. However, the first effects of the measures have already been noted, because a first estimate performed in September 2003 showed that nearly 90% of NGAP centers now have threedimensional dosimetry, 50% have a multileaf collimator, and 20 centers dispose of a dedicated CT scan. In addition, about 40 new authorizations for additional megavoltage units were delivered for both NGAP and DG centers, and the cancer plan has scheduled cobalt unit suppression for 2005. CONCLUSION Our study, done to improve the planning of RT in France, was a useful inventory. It showed that the technological RT infrastructures were not at the level associated with the modern practice of the specialty. Fundamental modifications in policy were essential for updating facilities and adequately covering the needs of the population. The first results are already visible. If updating equipment was a prerequisite, pouring in new resources only might not be sufficient to improve the quality of care. To progress in this way, recommendations for the good practice of RT, edited and, more importantly, advocated by professionals, are necessary. On this basis, other actions, such as quality assurance granted by contract between professionals and policymakers, could be implemented.

REFERENCES 1. Remontet L, Esteve J, Bouvier AM, et al. Cancer incidence and mortality in France over the period 1978 –2000. Rev Epidemiol Sante Publique 2003;51:3–30. 2. Kesteloot K, Lievens Y, van der Schueren E. Improved management of radiotherapy departments through accurate cost data. Radiother Oncol 2000;55:251–262. 3. Slotman BJ, Leer JWH. Infrastructure of radiotherapy in the Netherlands: Evaluation of prognoses and introduction of a new model for determining the needs. Radiother Oncol 2003; 66:345–349. 4. Kesteloot K, Pocceschi S, van der Schueren E. Reimbursement for radiotherapy treatment in the EU countries: How to encourage efficiency, quality and access? Radiother Oncol 1996;38:187–194. 5. Société Française de Radiothérapie Oncologie (SFRO) Livre Blanc. Evaluation et assurance qualité du plateau technique en Oncologie Radiothérapie. Bull Cancer Radiother 1996; 83(Suppl. 2):242S–272S.

6. Belletti S, Dutreix A, Garavaglia G, et al. Quality assurance in radiotherapy: The importance of medical physics staffing levels. Recommendations from an ESTRO/EFOMP Joint Task Group. Radiother Oncol 1996;41:89 –94. 7. Cortes F, Díaz R, García V, et al. Segundo Libro Blanco de la Oncología Española. Madrid: ENE Publicidad; 1995. 8. Inter-Society Council for Radiation Oncology. Radiation oncology in integrated cancer management. Philadelphia: InterSociety Council For Radiation Oncology; 1999. 9. Cancer. Le Plan de Mobilisation Nationale. Paris: Ministère de la santé, de la Famille et des Personnes handicapées; 2003. 10. Mackillop W, Fu H, Quirt C, et al. Waiting for radiotherapy in Ontario. Int J Radiat Oncol Biol Phys 1994;30:221–228. 11. Mackillop WJ, Groome PA, Zhang-Solomons J, et al. Does a centralized radiotherapy system provide adequate access to care? J Clin Oncol 1997;15:1261–1271. 12. Bernier J, Horiot JC, Bartelink J, et al. Profile of radiother-

516

I. J. Radiation Oncology

● Biology ● Physics

apy departments contributing to the Cooperative Group of Radiotherapy of European Organization for Research and Treatment of Cancer. Int J Radiat Oncol Biol Phys 1996;34:953–960. 13. Esco R, Palacios A, Pardo, et al. Infrastructure of radiotherapy in Spain: A minimal standard of radiotherapy resources. Int J Radiat Oncol Biol Phys 2003;56:319 –327. 14. van Daal WAJ, Bos MA. Infrastructure for radiotherapy in the

Volume 61, Number 2, 2005

Netherlands: Development from 1970 to 2010. Int J Radiat Oncol Biol Phys 1997;37:411– 415. 15. Teruki T, Owen JB, Hanks GE, et al. A comparison of the structure of radiation oncology in the United States and Japan. Int J Radiat Oncol Biol Phys 1996;34:235–242. 16. Berlan Y. Mission “Démographie des Professions de Santé”. Paris: Ministère de la santé, de la Famille et des Personnes handicapées; 2002.