JOURNAL TRANSCRIPT

---

Jpn J Clin OncoI1999;29(4 )209-213

National Average for the Process of Radiation Therapy in Japan by Patterns of Care Study Kazuaki Tanisada1, Teruki Teshima1, Toshihiko Inoue2, Jean B. Owen 3, Gerald E. Hanks4, Mitsuyuki Abe5, Hiroshi Ikeda6, Shinichiro Sato7, Kiyomitsu Kawachl/, Takashi Yamashita 8, Masamichi Nishio9, Masahiro Hiraoka10, Yutaka Hirokawa 11, Masahiko Oguchi 12 and Kouji Masuda13

Background: A nationwide effort is in progress to establish the actual state of radiotherapy and its quality assurance (QA) in Japan by using the Patterns of Care Study (PCS). In this study, national averages are calculated with a limited number of patients. A calculation program for national averages was prepared and applied to the radiotherapeutic processes used for esophageal cancer patients entered in the PCS. Methods: The calculation program for national averages, which were revised on the basis of differences between individual facilities and institutional strata, was developed in accordance with Sedransk's equation for the original PCS in the USA. National averages for several aspects concerning the sampled patients who had esophageal cancer between 1992 and 1994 were calculated with these procedures. Data for facilities and stratification of institution were simulated from a national structure survey of radiation oncology in 1990. Results: Values of the national average by Sedransk's equation were different from those of the simple sample average. There were significant differences in radiotherapeutic processes among stratification of institutions. For esophageal cancer, national averages were 0.129 for applications of endoscopic ultrasound, 0.599 for 'all fields treated each day' and 0.088 for application of brachytherapy. Conclusion: National averages for radiotherapy could be calculated. The values obtained in this PCS will be a useful measure for future QA in radiation oncology and in other specialties in Japan.

Keywords: Patterns of Care Study - national average - quality assurance- radiation therapy

INTRODUCTION Received 29 September 1998; accepted 28 December 1998 For reprint requests and all correspondence: Teruki Teshima, Department of Medical Engineering, Osaka University Medical School, 1-7 Yamadaoka, Suita, Osaka 565-0876, Japan. E-mail: [email protected] Abbreviations: PCS, Patterns of Care Study; QA, quality assurance; AI, university hospitals/cancer centers treating 300 patients or more per year; A2, university hospitals/cancer centers treating less than 300 patients per year; B1, other institutions treating 120 patients or more per year; B2, other institutions treating Jess than 120 patients per year; US, ultrasound

The Patterns of Care Study (PCS) for radiation therapy is a nationwide study by radiation oncologists to improve the quality and accessibility of radiation therapy in the USA (1). Because there has been no precedent of presenting the actual state of radiotherapeutic processes used in radiation oncology in Japan and of carrying out a nationwide clinical quality assurance (QA) study, the pes was imported from the USA. The data collection method ofPCS consists oftwo steps of random sampling (2). The first step is the classification of entire institutions into certain

nationwide strata and the random selection of institutions from

© 1999Foundation for Promotion of CancerResearch

Downloaded from https://academic.oup.com/jjco/article-abstract/29/4/209/795988 by guest on 28 November 2018

Departments of 1Medical Engineering and 2Radiation Oncology, Osaka University Medical School, Suita, Osaka, Japan, 3American College of Radiology, Philadelphia, PA, 4Department of Radiation Oncology, Fox Chase Cancer Center, Philadelphia, PA, USA, 5National Kyoto Hospital, Kyoto, 6Department of Radiation Oncology, National Cancer Center Hospital, Tokyo, 7Nationallnstitute of Radiological Sciences, Chiba, 8Department of Radiation Oncology, Cancer Institute, Tokyo, 9Department of Radiology, National Sapporo Hospital, Sapporo, 10Department of Therapeutic Radiology and Oncology, Kyoto University Graduate School of Medicine, Kyoto, 11 Department of Radiology, Hiroshima University Medical School, Hiroshima, 12Department of Radiology, Faculty of Medicine, Shinshu University, Matsumoto and 13Department of Radiology, Faculty of Medicine, Kyushu University, Fukuoka, Japan

National average for the processofradiation therapy in Japan

210

each stratum. The second step is the random selection of eligible patients from each sampled institution.

Y hik

.~:,.,.,.~~~~~

A

:

~

.

:

Figure 1. Con cept scheme of double random sampling in the Patterns of Care Study. L is numb er of strata in the whole nation ; h is the stratum of institution; there are Nil radioth erap y institutions in stratum h; nil is the number of radiotherapy institutions sampled in stratum h, indicated by the number of solid arrows ; (h, i) is the ith radiotherapy institution sampled in stratum h; there are Mh; patients treated in radiation institution (h, i); nw is the number of sam pled patients treated in institution (h, i) , indicated by the number of dotted arrows; Yilik is data for the ,zth sampled patient treated in facility (h, i).

If patient (h, i, k) is included in the calculation, Xhik is 1; otherwise Xhik is 0, and if Xhik is equal to 0, Yhik is O. This equation is similar to the simple method for calculating an arithmetic mean. If calculation of the national average uses randomly sampled data with adjustments of differences among institutional strata

METHODS AND MATERIALS CALCULATION OF NATIONAL AVERAGES USING SEDRANSK'S EQUATION

m~ \"'\~\~iTIT!11~iTi i!rIT

(3)

A

It was assumed that the entire radiotherapy facilities are divided into L strata and that there are Nh facilities in stratum h (Fig. 1). The ith facility (h, i) belonging to stratum h has Mhi patients and patient (h, i, k) is the kth patient in facility (h, i). Yhik is the value for patient (h, i , k). If Yhik is 1, this corresponds to ' Yes' and if Yhik is 0, this corresponds to 'No ' . If calculation of the national average, uses data for all patients, the calculation is represented by equation (1).

and individual facilities, the ratio estimator,

e, of e is

A

8

(2)

e,

e=

h =1 ;=1 k=1

=

sum of Y for all eligible patients number of all eligible patients

(1)

where nh is the number of sampled institutions in stratum h, mhi is the number of sampled patient records in institution (h, i), Yh; and l\;the means of Yhik and Xhik at all institutions and Yhik and Xhik represent the Yhik and Xhik of all sampled patients. The variance estimator,

1

(nil-I)

+ n~h [

V(e), is

6 nh

M2tn ( m h;

m hi )2]

1- M il;

S21l;

(3)

Downloaded from https://academic.oup.com/jjco/article-abstract/29/4/209/795988 by guest on 28 November 2018

In this study, the ratio estimator, 8 , as the national average was calculated using the equation devised by Sedransk and Sedransk (3). In this equation , the national averages were calculated by making statistical adjustments of differences among institutions and among institutional strata . By using national averages, the actual state of radiotherapy can be assessed quantitatively and these data can be used as a standard for QA in radiotherapy. The PCS for esophageal cancer and uterine cervical cancer between 1992 and 1994 in Japan has been carried out by our group since July 1996, with the support of the Ministry of Health and Welfare (8-27 and 8-29). Differences among patients' backgrounds, work-up studies and treatment methods were examined through analysis in terms of institutional stratification (4,5). However, this study produced only simple ratios for radiother apeutic processes, without making adjustments for differences among institutions. Based on these data, a calculation program for national averages was prepared. The calculation of national averages for several processes was piloted in preparation for QA of radiotherapy in Japan, which was started as a new study with a grant (10-15) from the Ministry of Health and Welfare (principal investigator: T.I.).

Jpn J Clin OncoI1999;29(4)

211

Table 1. Institutions audited and responsible persons and staff who collaborated inthis study

where

mhi

Id

hik

k=l

d hi = -m-hi

Responsible person and staff

Sakai Municipal Hospital Suita Municipal Hospital Sumitomo Hospital National Sapporo Hospital Aomori Prefectural Central Hospital Tohoku University Hospital

Sakai Suita Osaka Sapporo Aomori

Niigata University Hospital

Niigata

National Kyoto Hospital

Kyoto

Kyoto University Hospital

Kyoto

Kyushu University Hospital

Fukuoka

Hiroshima University Hospital

Hiroshima

Hamamatsu University Hospital

Hamamatsu

Shinshu University Hospital

Matsumoto

Gunma University Hospital

Maebashi

National Cancer Center Hospital

Tokyo

Tokai University Hospital

Isehara

Osaka University Hospital

Suita

Cancer Institute Hospital

Tokyo

National Osaka Hospital

Osaka

Osaka Rosai Hospital Osaka Red Cross Hospital Kansai Denryoku Hospital Kyoto City Hospital

Sakai Osaka Osaka Kyoto

Hiroshima Red Cross and Atomic Bomb Survivors Hospital Chuden Hospital Onomichi General Hospital Chugoku Rosai Hospital

Hiroshima Hiroshima Onomichi Kure

Hiroyasu Yoshioka Masayuki Sato Jun Ueda Masamichi Nishio Sadao Watanabe Yoshinao Abe Shogo Yamada Yoshihiro Takai Kunio Sakai Tadashi Sugita Mitsuyuki Abe Toru Shibata Masahiro Hiraoka Yasumasa Nishimura Yasushi Nagata Kouji Masuda Satoru Uehara Junichi Omagari Yutaka Hirokawa Yukio Akagi Masao Kaneko Tetsuo Nishimura Shusuke Sone Masahiko Oguchi Hideo Niibe Norio Mitsuhashi Michitaka Yamakawa Hiroshi Ikeda Yoshikazu Kagami Minako Sumi Tomoyuki Mori Yukio Oizumi Toshihiko Inoue Takehiro Inoue Takashi Yamashita Masahiko Furukawa Masao Kobayashi Hiroshi Igaki Masanori Mitomo Masatoshi Ohtani Isao Tsukaguchi Giro Todo Daizaburou Hamanaka Katsumi Hayakawa Mototsugu Koishi Masaki Mori Kazuki Kashimoto Akira Naito Tetsuji Kiso Katsuro Hanaguri

Nagano Red Cross Hospital

Nagano

Youichi Okazaki

Hokushin General Hospital

Nakano

Kiyonobu Ito

Sendai

LMhfXhi

L

xi = LN

Location

i=l

h

17=1

SIMULATION OF NATIONAL AVERAGE USING STRUCTURE SURVEY BY TSUNEMOTO (6) A calculation program was written by means of SAS 6.12 (7,8) according to the methodology mentioned above. National averages for the radiotherapeutic processes used for esophageal cancer patients who received radiotherapyin Japan between 1992 and 1994 were calculated. The items calculated were whether patients had undergone endoscopic ultrasound examination, whether all fields were irradiated each day and whether patients were being treated with brachytherapy. These were typical items that had apparent differencesby the stratificationof institutionand were important for radiotherapeutic processes represented by binary data. Patient data, Yhik, were obtained from the pes for esophageal cancer patients in Japan carried out by our group from July 1996 to February 1997. External audits were performed by the principal investigators (M.A. and H.I.) of the cancer research groups on the member institutions who agreed to official requests for this survey by the Ministry of Health and Welfare in Japan (8-27 and 8-29) and their affiliated hospitals, as shown in Table 1. The number of institutions audited was 29 and the fifth pes data format produced in the USA was used. The total number of esophageal cancer patients surveyed was 455. These included patients who received definitive, palliative, preoperative or postoperative radiotherapy. Patients who had distant metastasis or were diagnosed with any other cancer within 5 years before the start of treatment were excluded. Japanese institutions of radiation oncology were stratified into four categories according to the facility master survey by Tsunemoto (6). On the basis of this stratification, 39 institutions were classified as Al (university hospital/cancer center treating 2300 patients/year), 51 institutions as A2 (university hospital! cancer center treating